Robert Shaw’s “Gawdrehsteeoommehree”

Looking through some old octavos, I found a score edited by Robert Shaw, often noted as one of the most influential choral conductors of the twentieth century (mainly his work in popularizing the act of singing in a choir). This Shaw, unlike the more famous George Bernard Shaw, was not quite aware of the different sounds of the English language. Both Shaws worked to create a way of accurately depicting the English language’s sounds for a large audience. Bernard Shaw willed a contest to design an ideal alphabet for the English language (which was finished in the sixties and actually is quite attractive and now Unicode supported). He wanted the alphabet to distinguish every sound of the English language, which meant having about 40 to 48 letters.

Robert Shaw, however, created a second line of lyrics in some of his earlier choral octavos to try to unify the sounds of a choir. Let’s look at his version of the first verse of “God Rest You Merry Gentlemen”:

Gawdrehsteeoommehree jentuhlmmen, Lehtnuhthing eeoo dĭsmayee. Reemmembuhr Kraheestahwuhr Sayeeveeuhr oouh zbohwuhrnnawn Krismuhsdayee; Too sayeevuhsawl from Sayeetanspahwuhr hooehnnooee oohr gawnnuhstrayee Ooo taheedingzov kuhumfohuhrtandjoee, kumfohuhrtandjoee, Ooo taheedingzov kuhumfohuhrtandjoee.

If I translate this into IPA, what I get is this:

/ɡɒd rɛst i̯u mːɛɹiː ʤɛntəlmɛn lɛt nʌθɪŋ i̯uː dɪsmei riːmɛmbɚ kɹɑist ɑwɚ sejiviɚ uəz bowɚn ɒn kɹɪsməs dei tuː seiv əs ɒl fɹɑm seitænz pɑwɚ hu̯ɛn u̯iː wɚ ɡɒn əstrei oːu tɑidɪŋz ɑv kʌːmfɚt ænd ʤɔi/

It is supposedly US English, but possibly influenced by the cot-caught merger that would have been present in Shaw’s growing up years in northern California, he seems not to be able to hear the difference among /a/, /ɑ/, and /ɒ/. He puts God, and on, and gone on the rounded British vowel /ɒ/ instead of /ɑ/ (in his ears having the same vowel as the word all), tidings and Christ get /ɑ/ instead of /a/, and from gets /ɑ/.

He also conflates the approximants with vowels. For you, he has /i̯uː/ instead of /juː/. For when, he has /hu̯ɛn/ instead of /hwɛn/ or /ʍɛn/.

Most importantly, his choice of the British pronunciation of the short o (in the US, it is /ɑ/ and in the UK it is /ɒ/) probably had some influence on the modern conflation of most of the low back vowels as a single nondescript vowel.

I cannot imagine this kind of thinking started with him, but it certainly did not end with him, as it is still a serious issue today in choirs. I might compare it to Transatlantic English, which was the preferred accent of movies for most of the early to mid twentieth century. By the late twentieth century, it fell out of favor. The choral community just hasn’t caught up with that trend yet.

Shaw is often quoted as saying “When we sing, we don’t sing words, but sounds that sound like words.” While that might be partially true, it is still important to sing the right sounds so that the sounds sound like the right words.

Singing in English

This post comes as something of a second part to my earlier post, How to Sing an “Ah” Vowel in Choir, in which I try to explain the different sounds of English that often get lumped into one or two vowels. This post is more about the bigger picture of pronouncing the English language when singing.

From hearing several choirs in my YouTube and Spotify shuffles, I notice that when choirs from just about anywhere sing any language but English, they pay some respect to the phonetic inventories of the language. When we sing in German, we make a big deal, for instance, about differentiating [u] from [ʊ], making sure to pronounce the word und as /ʊnt/ and not /unt/.

Not so in English.

For some reason, singers in choirs are taught to whittle down the English vowel inventory to fewer and fewer possible vowels. Usually, it means that the director is asking for “taller” vowels, which I suppose really means migrating toward the sound of /ɑ/, also called “script a” or “single-story a.” What this kind of vagueness is doing is not only modifying the sound of English into an alien tongue when singing, but, through inconsistent application, is also causing serious vowel disharmony and bad tuning.

Here’s what’s happening: When singing in General American English, singers are being taught to sing

“eye” as [ɑɪ] instead of [aɪ],
“ow” as [ɑʊ] instead of [aʊ],
“ah” as [ɑ],
“uh” as [ɑ] instead of [ə],
“o” (“God”) as [ɒ] instead of [ɑ],
“ash” as [a] or [ɑ] instead of [æ],
“er” as [ɜ] instead of [ɚ],
“air” as [ɛ] instead of [ɛɹ],
“or” as [ɔ] instead of [ɔɹ], and
“ar” as [ɑ] instead of [ɑɹ].

Here is that last sentence read aloud, to hear the IPA vowels I’m talking about:

What this means is that for every replacement here, many choirs are singing both or several variants at the same time. Some of these replacements are coming from trying to make General American English sound more like Received Pronunciation. Some of this is coming from a distaste of the sounds of GA.

Here’s what that sounds like. Taking as a random example Eleanor Daley’s “In Remembrance,” from her Requiem, I grabbed two recordings, one from a professional choir in Canada, the other from a collegiate choir in the Southeast. Here is what the professional choir sounds like, with the words and a broad IPA transcription:

Do not stand at my grave and weep. I am not there. I do not sleep. I am a thousand winds that blow.

du nɑt stænd æt maɪ ɡrev ænd wip aɪ æm nɑt ðɛɹ aɪ du nɑt slip aɪ æm ə θaʊzənd wɪndz ðæt bloʊ

Here is a collegiate choir who is using a few more variants on these vowels, which I’ve separated with a backslash, which is non-standard IPA (but when does IPA ever have to transcribe simultaneous speakers?):

du nɑ\ɒt stæ\and æ\at ma\ɑɪ ɡre\ɛv æ\and wip a\ɑɪ æ\am nɑ\ɒt ðɛɹ a\ɑɪ du nɑ\ɒt slip ɑ\aɪ æ\am ə\ɑ θa\ɑʊzənd wɪndz ðæ\at bloʊ

Perhaps I’ve exaggerated a bit on how many vowels have variants in that recording, but my ears do sense those variants when singing in choirs. Were I singing this passage, though, I would prefer the vowels from the first transcription.

Can we as U.S. choirs agree to embrace General American English when singing American music? We have about fourteen or so vowels in our language, which we keep trying to modify to eight or so. Every vowel can tune if everyone is doing it together, even American r’s. If singers treat them as diphthongs (that is, waiting until the last moment to switch to the second sound), they sound just fine and quite understandable. As for the lone “er” sound as in “turn,” that sound can sound fine if people don’t tighten it too much.

My best advice for performing a work in GA English is to look up every word in the dictionary and not to stray from those vowels (I recommend the Oxford Advanced Learner’s Dictionary or even Wiktionary). High singers should modify for their instruments in their extreme registers, but other than that, variant vowels makes for bad music and wrong vowels make for nonsense. 

There is no reason to be ashamed of our language.

Problems with Choctaw Spelling

In George Aaron Broadwell’s A Choctaw Reference Grammar, one of the first topics he covers is the language’s orthography. He lists them as “Traditional orthography,” which is the one that developed in the mid-nineteenth century by missionaries, “Mississippi Choctaw orthography,” which came from the Mississippi Band of Choctaw Indians in the mid-1970s, “Modified traditional orthography,” which developed in the late twentieth century to differentiate long from short vowels and to improve consistency, and the “Choctaw Bible Translation Committee orthography,” which is currently being used by that committee in creating the first complete translation of the Christian Bible. The one that Broadwell chooses for his extensive grammar is the “modified traditional orthography,” which uses the digraphs sh and ch instead of š and č, and uses lh for ł.

The sentence he uses to show each one is from Matthew 19:14, “Suffer the little children to come unto me.” Here is this sentence in each orthography:

Traditional: Ʋlla chipunta yʋt ʋm ʋla hi a hʋsh im ahni.
Mississippi: Alla čipǫtayat amalahíyą hašimahni.
Modified: Alla’ chipotayat amalahiiya hashimahni.
Committee: Alla chipota yat am ala hí ya hash im ahni.

(Quick glossary: Alla’ = child, chipota = small, yat = nominative marker, am = first person singular agreement class III, al = come, ahii = irrealis, ya = different subject marker, hash = second person plural agreement class I, im = agreement class III, ahni = allow).

All four systems use Roman letters, but differentiate different things. In spoken Choctaw, vowel length is rather important, but Traditional, Mississippi, and Committee all only vaguely indicate vowel length by either an acute accent or a macron, or show nothing at all. Much of the written Choctaw that I’ve seen still differentiates vowel laxness, which tends to be inconsistently applied in spoken Choctaw. Nonetheless, many people still write child as ʋlla or vlla instead of alla. In the word for bread, pallaska’, for instance, many people still spell it palʋska, as it is spelled at the Oklahoma School of the Choctaw Language site. If you click the word on that site and listen, the letter ʋ is no more lax than the other letters a in the word.

Another difference that pops up occasionally is double consonants, which do make a difference in meaning occasionally. The word hattak (man), for instance, has been spelled “hatak,” and indeed is the root of the tribe name “Atakapa,” or “man-eater.” Just like in Italian, the double consonant is simply held a bit longer than a single consonant. This normally is not an issue in spelling until you get to a double ch or sh. One of the words for river is “hachcha,” which looks unnecessarily long (this by the way is in the word Atchafalaya, or “long river,” which runs through Louisiana). Haccha might make the eye read it as hakcha. Hačča is a good compromise here. Homma (red) is often rendered homa in early writings, or in the names Oklahoma (red people) and Coahoma (red panther).

The most obvious difference among these orthographies, though, is word spacing. Modified lumps all the particles together, while traditional separates each particle out, as we tend to do when romanizing Japanese so that we can more clearly see the word.

So here is my crazy idea.

Why not write Choctaw in a different writing system? Obviously there would be a steeper learning curve, but it would have a certain elegance to it that Roman letters fail to give well.

I looked around for an existing writing system that would do well writing Choctaw, with its few letters, nine vowels (three forms of three vowels: a, aa, a, i, ii, i, o, oo, o), confusing word separation questions, and so forth. The one that stuck out to me was the writing system used by the Shan people of Myanmar. It is a slightly simpler form of Burmese script that is remarkably clear, beautiful, and quite circular. I remember reading somewhere that the symbol of the circle has been historically quite important to the nation, so it drew me further to that idea.

Here is the system:


So, like the Shan language, the most common Choctaw syllable is a consonant followed by the vowel a. Thus, when a consonant stands alone, the a is assumed. Otherwise, another vowel mark is added as the table indicates. The symbol on each of those vowels in the table is the glottal stop. To indicate that a consonant ends the syllable, a c-looking diacritic is added above it. Double consonants are shown with a ၢ symbol. For instance, the word hattak would be ဂတၢၵ်.

In this system, there is no spacing between words at all; it flows just like speech. The main two punctuation marks are ၊ and ။, the comma and period/full stop.

Here’s the sentence from above:

Alla’ chipotayat amalahiiya hashimahni.

If anything, this was an enjoyable exercise. It isn’t as difficult to memorize as the Cherokee syllabary, but still narrows it down to syllables.

Here’s another short example:

Naa yoppahoosh binnililittook.
I sat down happily.

Here is one from the October 31, 2016, Lesson of the Day at the School of Choctaw Language:

Fʋni ʋlwʋsha ish ʋpa hinla ho?
Fani alwashaish apahilaho?
Can you eat fried squirrel?

This writing system might clear up some real headaches when trying to make sense of older spellings of words in the language.

How to Sing an “Ah” Vowel in Choir

Choir directors have often told me that the “ah” vowel is the most difficult one to tune. The discussion often ends right there, or begins with explanations of how to shape the mouth or vocal placement, or even showing an “ah” from a model.

The issue, however, is not that people have trouble getting their mouths around a particular vowel, but that U.S. ears have a great deal of difficulty distinguishing the different open back vowels. So what one winds up with after a degree in singing or choral conducting is this International Phonetic Alphabet vowel chart (note the question marks):

Vowel Chart as many Americans conceive of it. All the lower and right ones are questions.

When a choir is asked to sing on an “ah” vowel, often they conceive of a migrating point among those question marks. This is because even well-trained singers were never drilled on the sounds that each of those vowels in question makes. This post will try to fill in the gap a bit. Continue reading How to Sing an “Ah” Vowel in Choir

Singing Sweelinck in Dutch Latin

For another flavor of Latin, let’s look at how Latin was pronounced in Amsterdam in the early modern period.

Based on the information given in Harold Copeman’s Singing in Latin, here is an IPA transcription of “Hodie Christus natus est” (“Today Christ is born”) that Jan Pieterszoon Sweelinck set and had published in 1619. It is probably Sweelinck’s best known choral piece.

Hodie Christus natus est. Noe!
ˈhɔdie ˈkrɪstʏs ˈnaːtʰʏs ɛst nɔe

Salvator apparuit in terra canunt angeli
zɑlfaːtʰɔɾ əˈpaːɾʏɪtʰ ɪn ˈtʰɛɾə kaːnʏntʰ ˈɑŋxəɫi

lætantur archangeli exultant justi
leˈtʰantʰʏɾ ɑɾxˈaŋxəɫi ɛkˈzʏltʰantʰ ˈjʏsti

Dicentes, “Gloria in excelsis Deo! Alleluia!”
dɛiˈsentʰɛs ˈxɫɔɾia ɪn ɛkˈzɛɫzɪs ˈdeɔ ɑlelʏja

I have yet to hear a recording that tries it. I will bet it is much easier to sing the alleluias when the u is at the front of the mouth than in the Italian style that is universally used.

Pronouncing the Pater noster in Modern English Latin

The following transcription is based on Harold Copeman’s transcription of a reading given by John Field and Charles Low of the Westminster School, who both followed the rules of John Sargeaunt’s explanation of “Old Style” or Modern English Latin. I’ve also included his transcription of the Gloria Patri (that’s right, pronounced PAY-trigh). Enjoy.

Pater noster qui es in cælis,
[ˈpʰeɪtʰɚ ˈnɒstɚ kwaɪ ɛz ɪn ˈsiːlɪs]

sanctificetur nomen tuum:
[saŋktɪfɪˈsitʰɚ ˈnoʊmɛn ˈtʰjuʊm]

adveniat regnum tuum:
[adˈvɛniat ˈɹɛɡnəm ˈtʰjuʊm]

fiat voluntas tua
[ˈfaɪæt βɔˈlʊːntʰas ˈtʰjua]

sicut in cælo et in terra:
[ˈsaɪkʊt ɪn ˈsiːloʊ ɛtʰ ɪn ˈtʰɛɹei]

et dimite nobis debita nostra,
[ɛtʰ ˈdaɪmɪtʰi noʊbɪs ˈdibɪtʰa ˈnɒstʰɹa]

sicut et nos dimitimus debitoribus nostris:
[ˈsaɪkʊtʰ ɛtʰ nɒs ˈdaɪmɪtɪmʊs dibɪˈtʰɔɹɪbʊs ˈnɒstɹɪs]

et ne nos inducas in tentationem:
[etʰ ni nɒs ɪnˈdjukas ɪn tʰɛntʰaʃiˈoʊnɛm]

sed libera nos a malo.
[sɛd ˈlaɪbəɹeɪ nɒs eɪ ˈmeɪloʊ]

Gloria Patri

Gloria Patri et Filio et Spiritui Sancto
[ˈɡlɔːɾia ˈpeɪtraɪ ɛtʰ ˈfaɪlioʊ ɛtʰ ˈspaɪrɪtʰjuaɪ ˈsæŋktʰoʊ]

sicut erat in principio et nunc et semper
[ˈsaɪkʊtʰ ˈɛɹaːtʰ ɪn pɹɪnˈsaɪpioʊ ɛtʰ nʌŋk ɛtʰ ˈsɛmpɚ]

et in sæcula sæculorum. Amen.
[ɛtʰ ɪn ˈsiːkjula siːkjuˈlɔɹʊm ˈeɪmɛn]

Tallis’s O nata lux

This was a piece written in the late sixteenth century by Thomas Tallis. Hence, it uses late Renaissance English Latin. What follows is my transcription.

O nata lux de lumine,
[ɔ ˈnæːtʰa ljuks di ˈljumɪni]

Ieſu redemptor ſeculi,
[ˈʤizju rɛˈdɛmptɔr sɛkjuləːi]

dignare clemens ſupplicum
[dɪŋˈnæri ˈkʰlimɛnz ˈsjuplɪkʰʊm]

laudes preceſque ſumere,
[ˈlɑudɛz prɛˈsɛskwi ˈsjumɛri]

qui carne quondam contegi
[kwəi ˈkʰarni ˈkʰondam ˈkʰontɛʤəi]

dignatus es pro perditis,
[dɪŋˈnatʰʊz ɛs pro ˈpɛrdɪtʰɪs]

nos membra confer effici
[noz ˈmɛmbræ ˈkʰonfɛr ˈɛfɪsəi]

tui beati corporis.
[ˈtjui biˈætʰəi ˈkʰɔrpɔɾɪs]

Pronouncing Vivaldi’s Gloria

Vivaldi wrote his Gloria (RV 589, though the text is obviously the same with the other two Glorias) around the year 1715 in Venice. This means that the Latin used should be eighteenth-century Venetian Latin. What follows is a recording and my IPA transcription. It’s a little different that what you might think. Continue reading Pronouncing Vivaldi’s Gloria

Monteverdi Unfiltered: How to read Monteverdi part books

If you have found yourself singing a Monteverdi piece in your choir, perhaps you might want to unveil the mystery behind its original formats and not put your whole trust in a modern editor.

In my experience, reading from the original is not only better for the environment (fewer pages) and historically fun (you do feel a lot more connected to the music when you sing from the original notation), but it is actually a bit easier and less hectic than reading from modern scores.

The good news is that many of Monteverdi’s pieces are available online at the International Music Score Library Project in their original formats. When you go to his composer page, look under the Collections tab for them.

These parts can look a bit foreign at first if you’re used to seeing round notes vertically aligned with all the others parts in a modern score. These parts typically have no bar lines, use strangely shaped notes and rests, are somewhat crudely printed, and use odd time signatures. In the immortal words of the early music musicologist Jan Herlinger, “Everything was strange and difficult to read before you learned how to read it!”

If you have never tried early notation before, Monteverdi’s music is a great place to start, since the music is usually quite clear and easy. So let’s start from the beginning of a part to learn this system. Continue reading Monteverdi Unfiltered: How to read Monteverdi part books

Mozart’s Requiem in German Latin

The LSU Choirs were recently toying with the idea to perform the Mozart/Süßmayr Requiem in modern German Latin. By Modern, I mean the pronunciation as currently practiced by German choirs, which differs slightly from the Latin Mozart would have heard (for instance, Mozart would have heard [ˈʦɛli] rather than [ˈʦœli] and [ˈkʰiɾiɛ] instead of [ˈkʰyɾiɛ], as the moves to those vowels happened around 1850).

To aid in this effort, one of the choral conducting graduate students asked me if I would be willing to provide an IPA transcription of the words of the choir into German Latin. They are in this PDF and reproduced below. There is also here a recording of myself reading the text. I have not included the solo and quartet movements. Continue reading Mozart’s Requiem in German Latin