Loading ...
Sorry, an error occurred while loading the content.

Re: Pronunciation and writing of _r_ in Quenya

Expand Messages
  • Hans Georg Lundahl
    As far as I can see, unless my memory fails me - confining my knowledge to appendices of _LotR_ - the _rómen_ is an R as pronounced in _perro_ in Castille and
    Message 1 of 16 , Jan 5, 2004
    • 0 Attachment
      As far as I can see, unless my memory fails me - confining my knowledge to appendices of _LotR_ - the _rómen_ is an R as pronounced in _perro_ in Castille and the _óre_ an R as pronounced in _perro_ in some parts of South America - as the French pronounce J. Confirming this, Tolkien used _óre_ for English mute/Western R. I do not know if either was ever pronounced as R in _peró_ in Castille.

      Höstrusk och grå moln - köp en resa till solen på Yahoo! Resor

      [Non-text portions of this message have been removed]
    • Ales Bican
      ... **As I am very interested in these matters (indeed I have spent a lot of time thinking about Quenya phonetics and phonology recently), is it possible for
      Message 2 of 16 , Jan 5, 2004
      • 0 Attachment
        Helios De Rosario Martinez wrote:

        >I skip the discussion on the variants _rd_, _ry_, _hr_ and _rr_, and
        >go directly to the controversial item: the duality of the pronunciation and
        >spelling of the single _r_.

        **As I am very interested in these matters (indeed I have spent
        a lot of time thinking about Quenya phonetics and phonology
        recently), is it possible for you to share your insights with
        us (or me off-list if need be)?

        >We found the main information in the Appendix E:
        >
        >"R represents a trilled _r_ in all positions; the sound was not lost
        >before consonants (as in English _part_)." (LR:1088)

        **As far as I can say, these can mean basically two things: (1) by
        "trilled" Tolkien wanted to stress that Quenya _r_ is not mute and
        is pronounced in each and every position unlike English _r_ (as is
        after all suggested by the second part of the sentence) or, and which
        I find more likely, (2) Quenya _r_ is indeed trilled in all positions
        and (though this is not necessirily implied from the sentence) that
        Quenya _r_ is different to English _r_.

        >"[Óre] was often used for a weak (untrilled) _r_, originally occurring
        >in Quenya and regarded in the system of that language as the weakest
        >consonant of the _tincotéma_." (LR:1094)

        **Unlike Edouard Kloczko, I would read this in such a sense
        that Quenya (as distinguished from CE or PQ) had originally
        two types of _r_'s but later (in, say, Exilic Quenya) only one
        was prevalent.

        >"[Rómen] (in origin a modification of [óre]) was used for 'full'
        >trilled _r_" (LR:1095)

        [snip]

        >So, we may infer that when Fëanor invented the _tengwar_ there were
        >a weak untrilled and a full trilled _r_, clearly different, but in
        >later stages there was only one trilled phoneme.

        **I would like to note that it is not certain whether the
        original tengwar as devised by Feanor really had two
        characters for different _r_'s. What I want to say is that
        <rómen> could have been introduced by later loremasters.

        >The phonemes we are dealing with must be the trilled, tap and
        >approximant alveolars. If we take the literal meaning of the
        >citations, we learn that the phoneme the of _r_ in Third Age was the
        >alveolar trilled, as well as the original sound of _rómen_; and that
        >the original sound of _óre_ should be tap or approximant. We found the
        >approximant a likelier option, since the other _tengwar_ of the sixth
        >grade also represented Quenya approximants (_anna_ == /j/, and _wilya_
        >== /w/).

        **First of all, we do not know whether the rhotics (the _r_ sounds)
        you mention as possible (i.e. a trill, tap and approximant) are really
        alveolar. The contact point or place of approximation (in case of the
        approximant) may also be the dental region, which may be even more
        probable, because _r_ is listed as one of dentals when Tolkien
        mentioned which sounds were permitted word-finally (see Letters
        no. 347), though 'dental' in this case could be just a label that
        unites all these sounds.

        As regards the identity of Quenya _r_, I agree that we must think
        of only (voiced) dental/alveolar trill, tap/flap and approximant.
        All of them most likely apical (contact/approximation made with
        the tip of the tongue).

        I think it would be good to explain what trill, tap/flap and
        approximant rhotics are.

        Trills involve vibration of some articulator which is most often
        the tip of the tongue. The tongue tip is vibrating against a contact
        point in the dental/alveolar region. Trills usually consist of two
        or three periods of vibration.

        Taps (also called flaps, though some authors distinguish between
        taps and flaps) are those rhotic sounds that have only a single
        short closure. They typically involve a direct movement of the
        tongue tip to a contact location in the dental or alveolar region.

        Approximant _r_'s are those rhotics that do not involve contact
        between articulators but only approximation between them. In case
        of e.g. apical dental/alveolar approximant this means that the tip
        of the tongue makes approximation toward the dental/alveolar
        region but makes no contact. Note that this approximation is
        not accompanied by friction (friction is involved in fricative
        _r_'s).

        Now an apical trill occurs in Peninsular Spanish, Finnish and
        Italian (inter alia). A tap occurs in Spanish (usually described
        as dental) or in American English (a different type of tap -- if
        a distinction between tap and flap is maintained, the AE variety
        is rather a flap). Finally, an (alveolar) approximant rhotic is
        typical of Southern British English.

        Let me note that the above information is taken from a book
        called _The Sounds of the World's Languages_ by Peter
        Ladefoged and Ian Maddieson (1996). Another piece of
        information relevant for the present discussion is the fact that
        in Spanish a "weak" _r_ (i.e. the tap) occurs word-finally
        while a trilled _r_ (i.e. the trill) occurs word-initially and
        sometimes intervocalically where it sometimes stands in
        opposition with the tap. This information is taken (though also
        suggested in the aforementioned book) from _Fonetické
        obrazy hlásek_ (Phonetic images of sounds) by Bohuslav Hála
        (1960). Finally, I would like to mention that a book _Finnish_
        by Helen Sulkala and Merja Karjalainen (2002) confirms that
        Finnish _r_ is a trill (they speak about apico-alveolar trill)
        but the book also mentions that in word-medial position
        various allophones from one flap to several flaps appear.

        With this information at hand I can return to Quenya. My
        opinion is as follows: Tolkien spoke about a trilled _r_ and
        I think he really meant the (apical) dental/alveolar trill and
        not the Southern English approximant, because a trill appears
        in Finnish, Spanish and Italian. All of these languages Tolkien
        knew and liked. The untrilled variety Tolkien mentioned in
        App. E is also not (in my opinion) the English approximant
        but a tap/flap. I find it more likely because a tap occurs in
        Spanish (and also in Finnish). And there is also a striking
        resemblance in distribution of the sound. In Spanish a tap
        occurs word-finally; in Quenya the untrilled _r_ is represented
        by <óre> and this tengwa is usually used for word-final _r_'s.
        Furthermore, in Spanish a trill occurs word-initially; in Quenya
        the trilled _r_ is represented by <rómen> and this tengwa
        is usually used for word-initial _r_'s.

        >But others stated that the citations must be interpreted otherwise,
        >since they are oriented to a general (though interested) English
        >public, not to phoneticians, and that the terminology used in the
        >Appendixes does not accurately match the phonetic terminology we
        >are used to, say, the one of the IPA.

        **Appendix E is a problematic text. When reading it, we must
        remember that Tolkien was writing it primarily for English
        readers that could not be supposed to have deep linguistic
        knowledge. Also, he must have been limited by space. And
        he was not or could not be always accurate.

        For instance he says that _h_ in Quenya _eht_, _iht_ had the
        same sound as heard in English 'hew' or 'huge', and that _ht_
        had the sound of _cht_, as in German _echt_ (i.e. ich-Laut),
        _acht_ (i.e. ach-Laut). However, the sound in German _echt_
        is a voiceless palatal fricative but the sound in English
        'hew' is a voiceless palatal approximant (though the difference
        is subtle, of course).

        Another similar case is when Tolkien says that _h_ has the
        sound of English _h_ in 'house' or 'behold'. The problem is
        that the _h_ in 'house' is voiceless but voiced in 'behold'.

        Describing situation in Spanish Helios wrote:

        > - "'full' trilled" is like "vibrante múltiple" ("trilled" for IPA).
        > - "trilled" is like "vibrante" ("trilled" or "tap" for IPA).

        **Do you mean "untrilled" or "tap" for IPA?


        Ales Bican

        --
        What's in a name? That which we call a rose
        by any other name would smell as sweet. (Juliet, _Romeo and Juliet_)
      • Doug Pearson
        ... [snip] ... Not in American English: both _h_s are voiced and sound identical. It seems strange that Tolkien, (who I assume would differentiate these
        Message 3 of 16 , Jan 5, 2004
        • 0 Attachment
          Ales posted:

          > **Appendix E is a problematic text. When reading it, we must
          > remember that Tolkien was writing it primarily for English
          > readers that could not be supposed to have deep linguistic
          > knowledge. Also, he must have been limited by space. And
          > he was not or could not be always accurate.

          [snip]

          > Another similar case is when Tolkien says that _h_ has the
          > sound of English _h_ in 'house' or 'behold'. The problem is
          > that the _h_ in 'house' is voiceless but voiced in 'behold'.

          Not in "American" English: both _h_s are voiced and sound
          identical. It seems strange that Tolkien, (who I assume
          would differentiate these sounds) would make this mistake.
          Could he have had an American audience in mind?

          -- Tobold (Doug Pearson)

          [The answer to the final question is almost certainly "no". As
          for Tolkien having made a "mistake" -- it is far more likely
          that Tolkien was simply speaking of the pronunciation of
          _h_ in a general manner for a lay audience. I note that even
          the OED indicates the pronunciation of the _h_ in both 'house'
          and 'behold' with the same symbol 'h', which the "Key to the
          Pronunciation" indicates is pronounced as in 'ho!" -- PHW]
        • mach
          ... Why would Feanor create a letter for a sound he didn t use? Sure, he also created letters for aspirated sounds, but I guess this was rather because these
          Message 4 of 16 , Jan 6, 2004
          • 0 Attachment
            Helios cited from app. E:

            > > "[Óre] was often used for a weak (untrilled) _r_, originally occurring
            > > in Quenya and regarded in the system of that language as the weakest
            > > consonant of the _tincotéma_." (LR:1094)

            Edouard Kloczko commented:
            > I read here "originally occurring in Quenya" as meaning "originally
            > occurring in Common Eldarin (?)or/and Primitive Quendian", e.g. *not* in
            > Quenya as the living language of Aman and later in Exile, in which we had
            > only one trilled r.

            Why would Feanor create a letter for a sound he didn't use? Sure, he also
            created letters for aspirated sounds, but I guess this was rather because
            these sounds were observed in the language of the Valar. Or can we suppose
            that the lambengolmor of Feanor's time already knew the sounds of Common
            Eldarin (?)or/and Primitive Quendian? To my understanding, these sounds
            could only be reconstructed when the Eldar knew Sindarin and other Elvish
            dialects from Middle-Earth. And if we suppose that these sounds were known
            at the time of Feanor: Why would a letter for a sound that was only
            distinguished in Common Eldarin (?)or/and Primitive Quendian become part of
            the standard orthography of Quenya?

            As I understand it, there's only an external explanation for the duality of
            r-letters. J. R. R. Tolkien developed these two letters for a phonetic
            representation of his own "r-dropping" dialect of English: rómen for
            approximant _r_ and óre for dropped _r_, that is, for the schwa-sound that
            replaces an original _r_, e.g. in _here, there, under_ (it's difficult to
            decide whether óre is a vowel letter or a consonant letter).

            When he wrote Quenya with tengwar, he used both letters in almost the same
            way he used them in English. Maybe this was only because he was used to do
            so, or maybe because he liked to have diverse letters. Maybe Tolkien's
            attitude to the tengwar was somewhat lax, similar to Feanor's: "[...] in any
            case his primary interest was in _writing_, in its practical and its
            decorative aspects rather than in an accurate phonetic transcription. Not
            that he was with without interest in phonetic analysis" (app. E to _Quendi
            and Eldar_, in: VT 39, p. 8).

            Maybe his use of the two r-letters in Quenya remained unconscient until he
            had to explain how the tengwar work, that is, at last until he wrote the
            appendices to the Lord of the Rings. Helios' above quote might mean that J.
            R. R. Tolkien planned to revise the Quenya phonology by introducing a
            distinction of approximant _r_ vs. trilled _r_, but since his attitude to
            spelling matters was similar to Feanor's, he forgot to do so.

            Is the hypothesis of such a planned but never fulfilled revision plausible,
            I mean, do we have any evidence for similar cases?

            ---------------------------
            j. 'mach' wust
            http://machhezan.tripod.com
            ---------------------------

            [Edouard's reading of "originally occuring in Quenya" (with regard to untrilled
            _r_) as meaning "originally occurring in Common Eldarin (?)or/and Primitive
            Quendian" is, I think, unjustified and highly idiosyncratic. If one applies Occam's
            Razor (the simplest explanation is most likely to be correct) to this question,
            the situation regarding trilled R and weak R in Quenya seems rather obvious.
            Though R came to be pronounced as "a trilled _r_ in all positions" (LR:1088),
            there was originally also a weak R in Quenya (either Old Quenya, or the more
            conservative form of the language spoken in Valinor, as opposed to Exilic
            Quenya). Tolkien's use of the tengwar Rómen versus Óre in his transcription
            of "Namárie" gives us a clear indication of where these two sounds originally
            occurred -- trilled R (rómen) was usual at the beginning of words or
            intervocalically, and weak R (óre) was usual before consonants and at the
            end of words. -- PHW]
          • Helios De Rosario Martinez
            ... [snip] ... Of course. I will translate the text (I wrote it in Spanish) and then post it. But it will be in another post (it is quite long). ... Yes, so it
            Message 5 of 16 , Jan 6, 2004
            • 0 Attachment
              Ales Bican wrote:
              >
              > Helios De Rosario Martinez wrote:
              >
              > >I skip the discussion on the variants _rd_, _ry_, _hr_ and _rr_,

              [snip]

              > is it possible for you to share your insights with
              > us (or me off-list if need be)?
              >

              Of course. I will translate the text (I wrote it in Spanish) and then
              post it. But it will be in another post (it is quite long).

              > Another piece of
              > information relevant for the present discussion is the fact that
              > in Spanish a "weak" _r_ (i.e. the tap) occurs word-finally
              > while a trilled _r_ (i.e. the trill) occurs word-initially and
              > sometimes intervocalically where it sometimes stands in
              > opposition with the tap.

              Yes, so it is. The tap also occurs in contact (after or before) a
              consonant, although in those positions it depends on the individual
              pronunciation: some people (I myself, for instance) slightly trill
              ante- and post-consonantal _r_ depending on the velocity of speech
              (specially when speaking slowly or emphasizing the word). But the rule
              is:

              - Initial _r_ and intervocalic _rr_: trilled.
              - In other positions (final or adjacent to a consonant), and
              intervocalic _r_: tap.

              Note that intervocalic _rr_ is not longer (as Finnish); it only marks
              that it is trilled, opposite to intervocalic _r_.


              > The untrilled variety Tolkien mentioned in
              > App. E is also not (in my opinion) the English approximant
              > but a tap/flap. I find it more likely because a tap occurs in
              > Spanish (and also in Finnish).

              That is the only point were we disagree. The sounds in Finnish and
              Quenya are a good hint, of course. But note that LR:1094 says:

              "Grade 6 was most often used for the _weakest or 'semi-vocalic'_
              consonants of each series. (...) Thus [óre] was often used for a weak
              (untrilled) _r_, _originally_ occurring in Quenya and regarded in the
              system of that language as the weakest consonant of the tincotéma."
              [emphasis mine].

              I learn from those words that the original consonant for _óre_ was a
              "semi-vocalic" dental, as _anna_ was a "semi-vocalic" velar and
              _wilya_ was a "semi-vocalic" labio-velar. And I think that this
              "semi-vocalic" is what in the terminology of the IPA is called
              "approximant".

              This is not opposite to the fact that later Quenya lost this
              approximant sound, and retained only the Finnish-like trilled (and
              tap, maybe).


              > Describing situation in Spanish Helios wrote:
              >
              > > - "'full' trilled" is like "vibrante múltiple" ("trilled" for
              > > IPA).
              > > - "trilled" is like "vibrante" ("trilled" or "tap" for IPA).
              >
              > **Do you mean "untrilled" or "tap" for IPA?
              >

              No, I mean "either 'trilled' or 'tap' for IPA".
              _Vibrante_ is (in Spanish phonetical terminology) an ambiguous term,
              since it can be referred to either _vibrante múltiple_ (trilled) or
              _vibrante simple_ (tap). The common point is that both are opposite to
              the approximant, which is nearly "mute" for Spanish ears.

              What I meant is that maybe Tolkien used the word "trilled" in this
              way, not with its specific meaning for IPA, but opposite to the "weak"
              (which he specified as "untrilled") approximant.

              Helios
            • David Kiltz
              ... Laurence J. Krieg provided us with an accurate transcription of recordings (made in 1952) where J.R.R. Tolkien reads, inter alia, Namárie and A
              Message 6 of 16 , Jan 6, 2004
              • 0 Attachment
                On 05.01.2004, at 23:12, Ales Bican wrote:

                > With this information at hand I can return to Quenya. My
                > opinion is as follows: Tolkien spoke about a trilled _r_ and
                > I think he really meant the (apical) dental/alveolar trill and
                > not the Southern English approximant, because a trill appears
                > in Finnish, Spanish and Italian. All of these languages Tolkien
                > knew and liked. The untrilled variety Tolkien mentioned in
                > App. E is also not (in my opinion) the English approximant
                > but a tap/flap. I find it more likely because a tap occurs in
                > Spanish (and also in Finnish). And there is also a striking
                > resemblance in distribution of the sound. In Spanish a tap
                > occurs word-finally; in Quenya the untrilled _r_ is represented
                > by <óre> and this tengwa is usually used for word-final _r_'s.
                > Furthermore, in Spanish a trill occurs word-initially; in Quenya
                > the trilled _r_ is represented by <rómen> and this tengwa
                > is usually used for word-initial _r_'s.

                Laurence J. Krieg provided us with an accurate transcription of
                recordings (made in 1952) where J.R.R. Tolkien reads, inter alia,
                'Namárie' and 'A Elbereth Gilthoniel' [Jim Allan: An Introduction to
                Elvish p. 152ff.].

                From the transcription it can be seen that trilled and tapped 'r' are
                virtually interchangeable in Tolkien's pronunciation of Elvish. Tapped
                'r' is numerically prevalent whereas trilled 'r' is found throughout
                before dental (a common phenomenon).

                Hence I think it is possible that Quenya is thought to feature (or
                would 'naturally' have) positional variants of its 'r's, oscillating
                between trilled and tapped 'r'.

                However, Tolkien would mean both a tapped and a trilled 'r' when
                using the cover term 'trilled' in the Appendices.

                Although the distinction between tapped (actually a 'one-trill' r) and
                trilled 'r' e.g. in Spanish can be heard clearly, I think most people
                would characterize those 'r's as 'trilled' when speaking without a
                linguistic background or making only a rough statement. And
                indeed a trill is distinguished from a 'tap' or 'flap' just by the
                number of taps/flaps. That is, a trilled 'r' has many flaps.

                Hence, I think it most likely that the 'r' originally represented by
                <óre> would have been an approximant or fricative. As for the point of
                articulation, the listing in the tincotéma might be suggestive as Ales
                has said (i.e. dental/alveolar/post-alveolar). A guttural approximant
                or fricative might also be considered.

                -David Kiltz
              • Ales Bican
                I wrote that my opinion was that by untrilled _r_ Tolkien had meant a tap/flap (occurring in Spanish) and not an approximant (occurring in English). ...
                Message 7 of 16 , Jan 6, 2004
                • 0 Attachment
                  I wrote that my opinion was that by 'untrilled' _r_ Tolkien had
                  meant a tap/flap (occurring in Spanish) and not an approximant
                  (occurring in English).

                  Helios De Rosario Martinez wrote:

                  >That is the only point were we disagree. The sounds in Finnish and
                  >Quenya [read: 'Spanish' -- ab] are a good hint, of course. But note that
                  > LR:1094 says:
                  >
                  >"Grade 6 was most often used for the _weakest or 'semi-vocalic'_
                  >consonants of each series. (...) Thus [óre] was often used for a weak
                  >(untrilled) _r_, _originally_ occurring in Quenya and regarded in the
                  >system of that language as the weakest consonant of the tincotéma."
                  >[emphasis mine].

                  **This is an interesting note, I have not considered it until now.
                  Since <vala> belongs to "semi-vocalic" parmatéma/labials, it
                  speaks for my assumption that _v_ is a labial approximant rather
                  than voiced counterpart of _f_. At least _phonologically_
                  (inferring from its distribution and from the general phonological
                  system of Quenya) I am inclided to regard it as a labial
                  approximant and not labial (labio-dental) voiced spirant.

                  [Perhaps of relevance here is Tolkien's note in _The Shibboleth of
                  Fëanor_ that in early Quenya "the labial spirant _f_ was bilabial, and
                  so remains in Vanyarin" (VT41:7). CFH]

                  >I learn from those words that the original consonant for _óre_ was a
                  >"semi-vocalic" dental, as _anna_ was a "semi-vocalic" velar

                  **You mean "semi-vocalic" palatal, because no "semi-vocalic"
                  velar appears in Quenya.

                  > and _wilya_ was a "semi-vocalic" labio-velar. And I think that this
                  >"semi-vocalic" is what in the terminology of the IPA is called
                  >"approximant".

                  **This is certainly a good point. But again we must be cautious
                  with Tolkien's terms, for if he writes "semi-vocalic", it does not
                  have to mean that all these sounds are of the very same nature.
                  Similarly, if he writes "dentals" and lists _t, s, n, r, l_, it
                  does not mean that all these sounds are dentals. I would not rely
                  much on the fact that <óre> belongs to Grade 6, because it does
                  not actually state anything about how it was pronounced. Suppose
                  Quenya really had a tap and not an approximant, do you think that
                  Tolkien would have hesitated to use <óre> for this sound or do you
                  think he would have used another _tengwa_? Even though <óre> does
                  not have to be phonetically the best representation of the tap, it is
                  nevertheless a very apt choice, since <óre> belongs to the
                  _tincotéma_, which is the series that is used to represent all
                  "dental" sounds that appear word-finally except for _l_:
                  <tinco>, <thúle> (?), <númen> and <silme> (supposing it is a
                  modification of <thúle>).

                  >This is not opposite to the fact that later Quenya lost this
                  >approximant sound, and retained only the Finnish-like trilled (and
                  >tap, maybe).

                  **Nor is it oppositive to the fact that Quenya could have
                  lost the tap.

                  >_Vibrante_ is (in Spanish phonetical terminology) an ambiguous term,
                  >since it can be referred to either _vibrante múltiple_ (trilled) or
                  >_vibrante simple_ (tap). The common point is that both are opposite to
                  >the approximant, which is nearly "mute" for Spanish ears.
                  >
                  >What I meant is that maybe Tolkien used the word "trilled" in this
                  >way, not with its specific meaning for IPA, but opposite to the "weak"
                  >(which he specified as "untrilled") approximant.

                  **I see. But still I think it more likely that he meant the tap by
                  "untrilled". If we disregard the loose information Tolkien gives
                  about _r_ in Quenya in App. E, we are only left with indirect
                  evidence from real languages. So if I were to decide whether the
                  "untrilled" _r_ is a tap or an approximant, I would (and will)
                  choose the tap, because a tap occurs in Spanish and distribution
                  of the Spanish tap and trill is very reminiscent of distribution
                  of Quenya <óre> and <rómen> (i.e. "untrilled" and "trilled" _r_).
                  However, as I said it is only indirect evidence.


                  Ales Bican

                  --
                  What's in a name? That which we call a rose
                  by any other name would smell as sweet. (Juliet, _Romeo and Juliet_)
                • Ales Bican
                  To my note that Tolkien was not always accurate in App. E, e.g. when he stated that Quenya _h_ was pronounced as _h_ in English house (pronounced with a
                  Message 8 of 16 , Jan 6, 2004
                  • 0 Attachment
                    To my note that Tolkien was not always accurate in App. E, e.g.
                    when he stated that Quenya _h_ was pronounced as _h_ in
                    English 'house' (pronounced with a voiceless glottal fricative)
                    and 'behold' (pronounced with a voiced glottal fricative) several
                    people responded.

                    Doug Pearson wrote:

                    > Not in "American" English: both _h_s are voiced and sound identical.

                    **I am not a native speaker of English, so I have to basically rely
                    on what I read in books. Arthur J. Bronstein in _The Pronunciation
                    of American English_ (1960) notes that the _h_ sound is frequently
                    voiced in intervocalic position, which I think means that the _h_
                    sound is usually voiceless. As an example of voiced intervocalic
                    _h_ the authors gives, inter alia, 'behold'.

                    Eddin Najetovic wrote:

                    > As a matter of fact I am quite certain that in the English spoken in
                    > Britain the _h_'es are voiceless in every position, including
                    > those between vowels.

                    **I took the information about the voiced intervocalic _h_
                    in British Southern standard English from the two books
                    I mentioned last time, i.e. _The Sounds of the World's
                    languages_ by Peter Ladefoged and Ian Maddieson, and
                    _Fonetické obrazy hlásek_ by Bohuslav Hála. I remember
                    I read it also in one of Daniel Jones' books but since I
                    have already returned the book to library I do not remember
                    whether it was _An Outline of English Phonetics_ or _The
                    Pronunciation of English_ (I will check it if necessary).

                    Patrick Wynne wrote:

                    >I note that even the OED indicates the pronunciation of the _h_
                    >in both 'house' and 'behold' with the same symbol 'h', which the
                    >"Key to the Pronunciation" indicates is pronounced as in 'ho!"

                    **Yes, you are right. I have checked the OED. I have also
                    checked my _Oxford Advanced Learner's Dictionary_ (2000)
                    and here also both words are written using the same character.
                    In the section of the book 'Pronunciation and phonetic symbols'
                    the _h_ character is said to be pronounced as _h_ in 'hat'.
                    The pronunciation of 'hat' is given between two slashes.
                    This is what I find essential, because slashes are normally
                    used for phonological transcriptions. Pronunciation given like
                    this is rather confusing, because it is neither phonetic nor
                    phonemic, it is rather quasi- or pseudo-phonemic but of
                    course it depends on what is meant by 'phoneme' and
                    'phonemic' (generative phonologists, for instance, rejected
                    the notion of 'phoneme' as conceived by the Prague school).
                    Anyway, what I want to say is that if the pronunciation of both
                    'house' and 'behold' is given with one and the same symbol
                    _h_, it is rather to indicate that English has only one _h_
                    "phoneme" which may be pronounced differently according
                    to its position.

                    To return to Tolkien's statement. It is questionable what
                    Tolkien really meant. He need not have been familiar with the
                    fact that English had actually two variants of _h_. Or perhaps
                    he did not realize that the _h_ in 'behold' can be pronounced
                    differently than in 'house'. We might consider it then as an
                    error (even though it does not have to seem so always, we must
                    keep in mind that Tolkien was not omniscient and perfect). Yet
                    it also depends on how _h_ is actually pronounced in Quenya.
                    I suppose that _h_ in e.g. _halla_ "tall" is a voiceless
                    glottal spirant but what about its negation *_alahalla_
                    "not tall, short"? Perhaps the _h_ was voiced here just like
                    in English 'head' vs. 'ahead'.

                    [You may be quite certain that Tolkien was intimately aware of
                    all the allophonic variations in every major and most minor
                    dialects of English. CFH]


                    Ales Bican

                    ps. Another discrepancy in App. E I mentioned last time
                    was the problem of Quenya _hy_, which Tolkien described
                    as two similar yet different sounds (a voiceless palatal
                    approximant and voiceless palatal fricative). Eddin Najetovic
                    agreed with me pointing out that IPA does not even have a
                    symbol for the voiceless palatal approximant. Well, the IPA
                    does not have a lot of symbols it should have, unfortunately,
                    it should have been devised better. But this is not what I want to
                    say. What I want is to raise a question: so what is the Quenya _hy_,
                    a fricative or approximant? And this concerns also _hw_ -- a voiceless
                    labiovelar fricative or voiceless labiovelar approximant?

                    --
                    What's in a name? That which we call a rose
                    by any other name would smell as sweet. (Juliet, _Romeo and Juliet_)
                  • Arden R. Smith
                    ... Really? I see no examples of a voiceless palatal approximant there. In fact, the examples given in Tolkien s description of Quenya _hy_ in Appendix E are
                    Message 9 of 16 , Jan 7, 2004
                    • 0 Attachment
                      Ales Bican wrote:

                      >ps. Another discrepancy in App. E I mentioned last time
                      >was the problem of Quenya _hy_, which Tolkien described
                      >as two similar yet different sounds (a voiceless palatal
                      >approximant and voiceless palatal fricative).

                      Really? I see no examples of a voiceless palatal approximant there.
                      In fact, the examples given in Tolkien's description of Quenya _hy_
                      in Appendix E are merely a subset of the examples given for [ç]
                      ("Voiceless palatal central fricative") by Geoffrey K. Pullum and
                      William A. Ladusaw in their _Phonetic Symbol Guide_ (Chicago and
                      London: University of Chicago Press, 1986), p. 30:
                      "Illustrated by the initial segment of English _hue_ in some
                      pronunciations, by the final sound of German _ich_, and by the
                      initial segment of Japanese _hito_."

                      > Eddin Najetovic
                      >agreed with me pointing out that IPA does not even have a
                      >symbol for the voiceless palatal approximant. Well, the IPA
                      >does not have a lot of symbols it should have, unfortunately,
                      >it should have been devised better.

                      You speak as though the IPA was graven in stone long ago and is
                      therefore impervious to change. The IPA has been revised many times
                      in the course of its history, and today's IPA differs in many
                      respects from Paul Passy's original 1888 creation. For example, the
                      IPA didn't differentiate between the voiced palatal approximant and
                      the voiced palatal fricative until *1989*! There's certainly nothing
                      preventing the Association Phonétique International from adding a
                      symbol for the voiceless palatal approximant, should it be deemed
                      necessary. But in the 116-year history of the IPA, it has apparently
                      *not* been deemed necessary.

                      If you need to express the voiceless palatal approximant in the IPA,
                      however, it's already easy enough to do: use [j] with a little
                      circle under it (thus voiceless [j]). This is what Ian Maddieson
                      does in _Patterns of Sounds_ (Cambridge: Cambridge U. P., 1984), p.
                      245. Incidentally, the list that is given there of languages
                      containing that sound doesn't contain any likely models for Quenya:
                      Malagasy, Yao, Klamath, Otomi, Mazahua, Hopi, and Aleut.



                      Postscript: Anyone interested in the history of the IPA should check
                      out the following (in addition to the Pullum and Ladusaw book cited
                      above):

                      Robert William Albright, "The International Phonetic Alphabet: Its
                      Backgrounds and Development." _International Journal of American
                      Linguistics_ 24 (January 1958).

                      Michael K. C. MacMahon, "Phonetic Notation", in: Peter T. Daniels and
                      William Bright (eds.), _The World's Writing Systems_. New York and
                      Oxford: Oxford U. P., 1996, pp. 821-46.

                      Additionally, if you have access to a library with a complete run of
                      _Le maître phonétique_, the organ of the Association Phonétique
                      International, you can see the alphabet growing and changing right
                      before your eyes.

                      --
                      *********************************************************************
                      Arden R. Smith erilaz@...

                      Perilme metto aimaktur perperienta.
                      --Elvish proverb
                      *********************************************************************
                    • Ales Bican
                      I mentioned that Tolkien s description in App. E of Quenya _hy_ may be two-fold: it may be either an approximant or fricative ... **As j. mach wust already
                      Message 10 of 16 , Jan 8, 2004
                      • 0 Attachment
                        I mentioned that Tolkien's description in App. E of Quenya _hy_
                        may be two-fold: it may be either an approximant or fricative
                        according to his description (in my view). Arden R. Smith wrote:

                        >Really? I see no examples of a voiceless palatal approximant there.

                        **As j. 'mach' wust already mentioned, I meant the sound in English
                        'hue' to be a voiceless palatal approximant.

                        >In fact, the examples given in Tolkien's description of Quenya _hy_
                        >in Appendix E are merely a subset of the examples given for [ç]
                        >("Voiceless palatal central fricative") by Geoffrey K. Pullum and
                        >William A. Ladusaw in their _Phonetic Symbol Guide_ (Chicago and
                        >London: University of Chicago Press, 1986), p. 30:
                        >"Illustrated by the initial segment of English _hue_ in some
                        >pronunciations, by the final sound of German _ich_, and by the
                        >initial segment of Japanese _hito_."

                        **According to Ian Maddieson and Peter Ladefoged (_The
                        Sounds of the World's Languages_ (2002, first published in
                        1996) Tolkien described two different sounds: "The onset
                        in [English _hue_] is normally a voiceless palatal approximant,
                        _j_ [written with an underposed circle], for which the IPA
                        has no unitary symbol." (326).

                        I wrote:

                        >>Eddin Najetovic
                        >>agreed with me pointing out that IPA does not even have a
                        >>symbol for the voiceless palatal approximant. Well, the IPA
                        >>does not have a lot of symbols it should have, unfortunately,
                        >>it should have been devised better.

                        Arden:

                        >You speak as though the IPA was graven in stone long ago and is
                        >therefore impervious to change.

                        **Well, I suppose this may be a reading of what I wrote. Next time
                        I should definitely attempt to use better words and phases. -- Of
                        course, I know that the IPA underwent a number of revisions, so
                        basically I wanted to say is that it still needs some revisions
                        to be better.

                        > The IPA has been revised many times
                        >in the course of its history, and today's IPA differs in many
                        >respects from Paul Passy's original 1888 creation. For example, the
                        >IPA didn't differentiate between the voiced palatal approximant and
                        >the voiced palatal fricative until *1989*!

                        **Revisions of the chart are (I suppose) based on current research,
                        so what was not (could not) be differentiated in the past can be
                        differentiated today. I do not know the state of affairs in Tolkien's
                        time but maybe it was not known then that the sounds of Eng.
                        _hue_ and Ger. _ich_ are different and therefore Tolkien did not
                        make any difference between them.

                        > There's certainly nothing
                        >preventing the Association Phonétique International from adding a
                        >symbol for the voiceless palatal approximant, should it be deemed
                        >necessary. But in the 116-year history of the IPA, it has apparently
                        >*not* been deemed necessary.

                        **The IPA is (as far as I can see) good for a language like English
                        but as far as I know Czech phoneticians do not perceive as
                        fitting for Czech, but of course it is also a matter of getting
                        used to it. At any rate a character for r-hacek (and no, it is _not_
                        [r] + [zh] (i.e. the sound in 'pleasure')!) is missing (why?). Also,
                        the way afficates are written in the IPA does not look very pleasing
                        to a Czech eye. Another thing that could be bettered is difference
                        between _þ, ð_ and _s, z_.The former are given dental fricatives
                        and the latter as alveolar fricatives. A distinction should, however,
                        be rather made between sibilant and non-sibilant fricatives, as we
                        can have both dental and alveolar _þ, ð_ and dental and alveolar
                        _s, z_.

                        >If you need to express the voiceless palatal approximant in the IPA,
                        >however, it's already easy enough to do: use [j] with a little
                        >circle under it (thus voiceless [j]).

                        **I know that but I and Eddin (I think) meant that the IPA did not have
                        a unitary symbol for it.

                        > This is what Ian Maddieson
                        >does in _Patterns of Sounds_ (Cambridge: Cambridge U. P., 1984), p.
                        >245. Incidentally, the list that is given there of languages
                        >containing that sound doesn't contain any likely models for Quenya:
                        >Malagasy, Yao, Klamath, Otomi, Mazahua, Hopi, and Aleut.

                        **I do not have access to the book, only to the book I mentioned
                        above, which is from 1996. Since this is newer book, I (not having
                        means to find out by myself) prefer to rely on this one. So if
                        English is said to possess this sound, it is a rather likely model.
                        Personally, however, I am inclined to the fricative (in German _ich_)
                        -- _hy_ (at least phonologically) does not appear as the voiceless
                        counterpart of _y_ in Quenya.

                        By the way, could you give me (us) a list of some languages that are
                        mentioned in "your" book? "My" book only mentions (with a reference
                        to "your" book) that less than 5% of the world's languages include
                        the voiceless palatal fricative in their inventory.


                        Ales Bican

                        --
                        What's in a name? That which we call a rose
                        by any other name would smell as sweet. (Juliet, _Romeo and Juliet_)
                      • Arden R. Smith
                        ... Then there s disagreement among phoneticians as to what sound the initial segment of _hue_ really is. I ve already mentioned Pullum and Ladusaw s
                        Message 11 of 16 , Jan 10, 2004
                        • 0 Attachment
                          Ales Bican wrote:

                          >**According to Ian Maddieson and Peter Ladefoged (_The
                          >Sounds of the World's Languages_ (2002, first published in
                          >1996) Tolkien described two different sounds: "The onset
                          >in [English _hue_] is normally a voiceless palatal approximant,
                          >_j_ [written with an underposed circle], for which the IPA
                          >has no unitary symbol." (326).

                          Then there's disagreement among phoneticians as to what sound the
                          initial segment of _hue_ really is. I've already mentioned Pullum
                          and Ladusaw's statement from 1986. Furthermore, William G. Moulton
                          says the following in his _The Sounds of English and German_ (Chicago
                          and London: University of Chicago Press, 1962), p. 29:

                          "The initial /h-/ of English /'hju/ _hue_, _Hugh_ [...] is pronounced
                          by many Americans as a [ç] with rather wide opening."

                          My point here is this: If *phoneticians* can't agree what sound
                          appears at the beginning of _hue_, then the distinction between these
                          two sounds must be so minuscule that it is of no practical
                          consequence. So you can go ahead and pronounce _hyarmen_ with a
                          voiceless palatal approximant, and I'll go on pronouncing it with a
                          voiceless palatal fricative, and no one will notice the difference.

                          >By the way, could you give me (us) a list of some languages that are
                          >mentioned in "your" book? "My" book only mentions (with a reference
                          >to "your" book) that less than 5% of the world's languages include
                          >the voiceless palatal fricative in their inventory.

                          I'm guessing that even fewer languages have a voiceless palatal approximant.

                          _Patterns of Sounds_ (p. 231) lists the following languages as having
                          the voiceless palatal fricative: Irish, Norwegian, Bengali, Komi,
                          Margi, Mandarin, Kan, Haida, Kwakw'ala, Paez, and possibly Chuvash.
                          Two things must be borne in mind: (1) This book discusses the
                          *phonologies* of the various languages, so only phonemes are treated
                          here, hence the absence of German (in which Maddieson regards [ç] as
                          an allophone of /x/); (2) The lists are by no means exhaustive (for
                          example, the only Germanic languages included are German and
                          Norwegian).

                          --
                          *********************************************************************
                          Arden R. Smith erilaz@...

                          Perilme metto aimaktur perperienta.
                          --Elvish proverb
                          *********************************************************************
                        • Andreas Johansson
                          ... That, or pronunciation varies from speaker to speaker and/or location to location. It s the kind of minor phonetic variation one expect to see abundantly
                          Message 12 of 16 , Jan 11, 2004
                          • 0 Attachment
                            Quoting "Arden R. Smith" <erilaz@...>:

                            > My point here is this: If *phoneticians* can't agree what sound
                            > appears at the beginning of _hue_, then the distinction between these
                            > two sounds must be so minuscule that it is of no practical
                            > consequence.

                            That, or pronunciation varies from speaker to speaker and/or location to
                            location. It's the kind of minor phonetic variation one expect to see
                            abundantly in a language of a few hundred million speakers.

                            Andreas
                          • David Kiltz
                            On 11.01.2004, at 06:20, Arden R. Smith wrote: A.R.Smith says that William G. Moulton says in his _The Sounds of English and German_ (Chicago and London:
                            Message 13 of 16 , Jan 11, 2004
                            • 0 Attachment
                              On 11.01.2004, at 06:20, Arden R. Smith wrote:

                              A.R.Smith says that William G. Moulton
                              says in his _The Sounds of English and German_ (Chicago
                              and London: University of Chicago Press, 1962 that Modern Standard
                              German [ç] sounds like the initial sound of _hue_ as pronounced by
                              many Americans. Now there definitely is a difference between the RP
                              _hue_ sound and the German _ich_ sound (_hue_ is slightly more forward)
                              but it's a miniscule difference. Note the fact that English /ç/
                              only occurs before /u/ word initially whereas German /ç/ never does.
                              That may also add to the difference, thus, perhaps, making the
                              difference between the sounds hardly more than a positional one. Still,
                              all Germans I asked (including me) notice and make a clear distinction
                              between English /ç/ and German /ç/ (and it is lab verifiable!).
                              Still, I think it fair to settle on /ç/ for Quenya as Tolkien's
                              examples can be approximations at times.

                              > Maddieson regards [ç] as
                              > an allophone of /x/ (in German).

                              Just for the record: Not only the vocalic environment plays a role here
                              but also morphemes. Cf. _Frauchen_ /frauçen/ vs. _rauchen_ /rauxen/.

                              -David Kiltz
                            Your message has been successfully submitted and would be delivered to recipients shortly.