Proposed System of Auxiliary (Diacritic) Markers to Aid Children and Adult Learners in Mastering Cherokee Script:

Preliminary Remarks:

Paper:

1. Voicing/Aspiration: The difference in pronunciation between the English words 'pear' and 'bear', considered by most people to be one of voicing, can be attributed as much to aspiration, I believe. Of course, there are minimal pairs like 'breech'/'preach' in which aspiration is said to play no role. Nevertheless, I tend to think of voicing and aspiration as auxiliary components of one and the same distinction, at least in English. I am told, however, that the difference in Cherokee is really more of aspiration than of voicing. In order to reduce potential confusion to those unfamiliar with the difference, however, I have decided to use both terms synonymosly in this paper. If you feel lost, all I mean by this is that when I'm contrasting similar sound pairs like 'd'/'t' and 'b'/'p' and 'g'/'k', I'm going to refer to the left element as 'voiced/unaspirated' and the right element as 'voiceless/aspirated'. If you still don't get it, don't worry. I'll explain in simpler terms in specific cases as they arise below.
   Consider the sequence of characters which, depending on the pronunciation of the first consonant, can mean either "winter" (GO-LA) or "bone" (KO-LA). Sequoyah was obviously aware of this distinction because he created separate symbols for five such pairs, these being (GA) and (KA), (DA) and (TA), (DE) and (TE), (DI) and (TI), and (DLA) and (TLA). I presume he would have created distinctions for all the two or three dozen pairs that occur in Cherokee (depending on dialect and, perhaps, other factors) except that this would have made the already somewhat cumbersome 85-character set too large for practical usage. Instead, he presumably chose to include only the most commonly occuring ones.
   
   To those who know Cherokee well, it is clear whether the meaning 'bone' or 'winter' is intended, I'm sure. But what about those of us who are new to both the script and the language? Even if we can guess at the meaning through context, how will we ever learn the correct pronunciation? Clearly the answer is not to create more characters. Particularly nowadays when there are so many people who are used to the system the way it is and with numerous books and articles in print which already use the existing system.
   
   Instead, I propose that, in cases like that of KO-LA/GO-LA above where there are potential problems resulting from ambiguity in terms of either pronunciation or meaning, a sign be placed above the character in question, explicitly marking it as voiced/unaspirated or voiceless/aspirated. I further propose that this sign be derived from the horizontal line and downward hook appearing at the top of the Cherokee character (TI) which is what differentiates that character from its voiced/unaspirated equivalent, (DI). Note also that the same stroke is found in the voiceless/aspirated syllable (TSE) as well as in voiceless/aspirated (TE). (This is clearer in the woodcut font or in the handwritten characters). I believe its occurence in these three characters makes it an appropriate candidate for a marker of voicelessness/aspiratedness. And I believe it is worthwhile to draw such symbols, insofar as possible, from the syllabary itself in order to preserve its unique character.
   
   The above proposal yields something like the following:
   
   

   Ambiguous Notation:				Explicit Notation:
   (Implies voicing but does	GO	LA		(Rules out voicing)	KO	LA
   not rule out devoicing)	KO	LA

   
   
   I further propose that this sign be inverted in order to explicitly notate voicedness/unaspiratedness, as well:
   
   

   Explicit Notation:
   (Rules out devoicing)	GO	LA

   
   
   Consider how these symbols could aid the learner in pronouncing the two words below, each of which contains both voiced and voiceless versions of an otherwise identical consonantal sound:
   
   

   A-TSU-JA = 'BOY':					U-GU-KU = 'OWL':

   
   
   
2. Vowel-length: Perhaps the most famous (infamous?) example of this type of ambiguity is that represented in the words for 'water' and 'salt', both spelled (A-MA) in the syllabary. I have this recurring nightmare that I'm wandering out in the desert somewhere in Oklahoma without water, quickly expiring of thirst. I come across a beautiful Cherokee woman and ask her for water. Only my pronunciation is so poor that she thinks I'm asking for salt. She lifts it up to my mouth and begins to pour. When it touches my lips, I wretch up the last little bit of moisture in my system and crumble into dust at her feet.
   
   But all joking aside, here's a simple idea for notating short and long vowels² in Cherokee. Simply place one horizontal line underneath a character containing a short vowel. For long vowels, use two lines, one on top of the other, as follows:
   
   

   Default notation:			Explicit notation:			Explicit notation:
   (Ambigious)			(Long Vowel)			(Short Vowel)
   'WATER' OR 'SALT'	A	MA	'WATER'	A:	MA	'SALT'	A	MA

   
   
   Of course, vowel length can often be zero, as well. Take, for example, the word 'SVKTA' meaning "apple" which is actually spelled (SV-GA-TA). To represent the zero vowel in the second character, we'll borrow a diacritic marker from Hebrew which is used to notate the absence of a vowel. That sign resembles a colon and is placed directly under the character in question. We'll place it under the appropriate Cherokee character in similar fashion. Incidentally, I don't know why the voiced/unaspirated syllable (GA) is used here instead of voiceless/aspirated syllable (KA), but no matter. We'll simply use the voiceless sign from section one above in conjunction with the zero-vowel symbol in order to modify the syllable 'GA', thus rendering it as 'K':
   
   

   SV-K-TA = 'APPLE':

   
   
   Two additional examples exist in the Cherokee words for "button" and "bug" which, according to Durbin Feeling's Cherokee Dictionary, are spelled (GA-TI-DI) and (TSI-GO-YA) but are pronounced GATDI and TSGOYA respectively. Using the zero-vowel marker, we can explicitly notate the correct pronunciations as follows:
   
   

   GA-T-DI = 'BUTTON':					TS-GO-YA = 'BUG':

   
   
   
3. Nasality: According to the textbook Beginning Cherokee by Ruth Bradley Holmes and Betty Sharp Smith, many Cherokee words are nasalized in the final syllable. They give the example (TSI-YI) meaning "copper", which is pronounced like this. (Click on the word "this" in the previous sentence to hear the word pronounced via a .WAV file). Incidentally, nasalized vowels are common in many other languages, as well, among them Chinese, French, Lakota, and Portuguese as well as Iroquoian languages like Mohawk which are spoken in and around New York State and which are close relatives to Cherokee.
   
   I propose that we use the small hook in the right top corner of (NV) to represent this sound. This seems appropriate because it is the only symbol representing both a nasalized consonant and a nasalized vowel. (It also sort of looks like an 'N', come to think of it... Well, at least in this version as hand-drawn by Sequoyah himself: ). Here's how the word for "copper" would look using that resulting symbol:
   
   

   TSI-YI = 'COPPER':

   
   
   In certain cases, you may want to show vowel length and nasality for/in the same syllable. In that case, you could use a combination sign that contains both elements as I have done below for both a long and a short vowel, respectively:
   
   

   VV-V = 'YES':				TSI-YI = 'COPPER':

   
   Of course, the nasality marker is superflous in the first example directly above since the symbol (V) is already nasal by definition. But until I can find a better example, I'll just have to keep using this one.
   
   
4. Intrusive 'h'/Glottal stop: The Cherokee word for "wolf", spelled (WA-YA), is pronounced with what some linguists refer to as an intrusive 'H'. This is an unwritten 'H' sound which intervenes between a previous syllable and the beginning consonant of the current syllable. The combination of the 'H' and 'Y' in WAHYA thus derived is similar to the initial sound in the English word "HUGE". If you were to phonetically spell the word "HUGE" using a hodgepodge of weird English spelling conventions, you might come up with something like HYOOJ. Take the "HY" portion of that spelling and place it into the middle of the Cherokee word WAHYA and, voilą! You have the correct pronunciation of the word meaning "wolf." This word is usually Romanized as "WAHYA" to reflect the extra sound not captured by the syllabary. For our purposes, I'll divide the syllables with a dash so as to ensure that the 'H' sound does not get ignored: WA-HYA
   
   The same thing happens with the glottal stop which occurs in such common Cherokee expressions as (WA'DO = Thank you). The glottal stop also occurs in English where it is likewise not written. A good example is the expression commonly spelled UH-OH which is pronounced with a catch in the throat at the beginning of both syllables. That catch in the throat is what we mean by 'glottal stop'. It occurs very frequently in Arabic and Farsi, and it is sometimes Romanized as an apostrophe in such names when they are transliterated into Roman script. One example of this is RA'ID, the name of a Jordanian prince whom I met once or twice while attending the University of Amman in 1994.
   
   I propose that the Greek 'rough breathing' mark be adopted here to represent intrustive 'H'. For the glottal stop, I originally proposed the Greek 'soft breathing' mark which is simply an inverted rough breathing mark. In order to help readers better distinguish the two symbols, Dr. Bender suggested I replace this with the dotless question mark, a symbol that is often used in Romanized transcriptions of Cherokee and other languages to denote this very sound and, thus, a symbol that many will already by familiar with.
   
   For the Greeks, "soft breathing" amounts to that catch in the throat which is what we identified in the previous paragraph as the glottal stop while "rough breathing" is nothing more than that aspiration we know as 'H'. The added tail will help to distinguish these two characters. My thanks to Margaret Bender for suggesting this.
   
   Here's how these symbols would like in the words (WAHYA) and (WA'DO):
   
   

   WA-HYA = 'WOLF':					WA'-DO = 'THANK YOU':

   
   Once again, these signs can co-occur with the vowel-length signs, as well:
   
   

   WA-HYA = 'WOLF':					WA'-DO = 'THANK YOU':

   
   By the way, you may wonder why I've chosen to place the sign for the intrusive 'H' at the left edge of the syllable while the sign for the glottal stop occurs at the right edge. This is because intrusive 'H' is pronounced as a part of the syllable which it follows while the glottal stop is a part of the previous syllable except when it occurs between two vowels. But there's no need to get too technical about special cases. Therefore, although there is a situation we'll discuss later on where a leftward glottal-stop symbol is used, I would generally use the rightward one. You could use the forward-positioned glottal marker for glottal stops which occur at the beginning of a word, I suppose, but glottal stops occuring in this position aren't usually all that important, anyway. It's kind of like saying "apple" with or without a glottal stop in English. It doesn't make any difference at all in terms of meaning.
   
   Keep in mind that this system is merely for disambiguating the pronunciation of particularly troublesome syllables. I don't envision diacritical marks on every syllable. And I don't know that it would ever be something that would make it into print, except in dictionaries or children's readers. Mostly, I can see these marks being scribbled into the pages of a personal hymnbook or Bible in pencil to aid in getting the right pronunciation when singing or reading aloud.
   
   
5. Pitch/Accent: While Cherokee is not a tonal language like Chinese or Navajo, linguists claim that pitch is phonemic in certain instances. For example, Sharp and Smith state that the difference between the words 'cold', 'north', and 'Republican' is one of pitch. (The same is true for 'warm', 'south', and 'Democrat', by the way). Four levels of pitch are identified, and these are generally marked with a superscript digit following the relevant syllable:
   
   U¹HYV²DLA 'cold'
   In some cases, the pitch rises or falls within the syllable, in which case two digits appear to show the pitch at the starting and ending point:
   
   DI²TE²³SDI 'iron'
   I recommend that pitch be notated with short vertical lines emanating up and down from the long horizontal line(s) which denote vowel length as in the following examples:
   
   

   U1HYV2DLA = 'COLD':						DI2TE23SDI = 'IRON':

Additional Issues:

• Consonantal alterations: In addition to the five types of ambiguity discussed above, there is another phenomenon that should be addressed here. In Cherokee, as in most other languages, sounds sometimes become altered in the context of surrounding sounds. For example, in at least one Cherokee dialect, the word (WA-LO-SI) meaning "frog" is pronounced as WALOSH. What happens here is that the /s/ sound becomes changed to the higher /sh/ sound (as in "shine") due to the influence of the high vowel (/i/) directly following it. (In other words, the quality of highness bleeds from /i/ into the surrounding environment just as purple dye might bleed from a pair of Donny Osmond's socks improperly placed into a batch of white undergarments). In the process, the vowel disappears leaving us with WALOSH (rhyming with gauche).
  
  I propose that a "consonantal alteration" symbol be created from (the sign for voicelessness) by adding a second downward stroke directly at the center of the horizontal line like this:
  Using this new symbol in conjunction with the zero-vowel symbol, we could then render a phonetic spelling of the word for "frog" as follows:
  
  

  WA-LO-SH = 'FROG':

  
  This convention could be applied to the sign for voicedness, as well, making it possible to notate other voiced consonants not covered by the syllabary. These symbols would then allow us to write virtually any English sound in the Cherokee script - something that was impossible before. This may not seem particularly useful, but it could prove beneficial in the future should another indigenous group such as the Mohawks or the Tohono O'odham consider adopting the only script indigenous to the United States in writing their own languages. I mention the Tohono O'odham in particular because there is, at present, a small literary movement flourishing there under the tutelage of poet and professor Ofelia Zepeda and others. I mention the Mohawk because they are known for their vigorous use of their indigenous language which is a close relative to Cherokee itself. Here are a few examples of how foreign sounds (ie. sounds not native to Cherokee) could be notated with the help of diacritic marks:
  
  
  • How items on the 'M' row could be altered to reflect other bilabial stops:
    
    

    MA	BA	PA		ME	BE	PE		MI	BI	PI

    
    

    MO	BO	PO		MU	BU	PU

    
    
  • How items on the 'W' row could be altered to reflect other bilabial fricatives:
    
    

    WA	VA	FA		WE	VE	FE		WI	VI	FI

    
    

    WO	VO	FO		WU	VU	FU		WV	VV	FV

    
    
  • How items on the 'D/T' row could be altered to notate the voiceless English 'TH' in "THEATRE" and the voiced English 'TH' in "THE"/"THEN" :
    
    

    DA

    TA

    THA

    THA

    DE

    TE

    THE

    THE

    DI

    TI

    THI

    THI

    
    

    DO

    TO

    THO

    THO

    DU

    TU

    THU

    THU

    DV

    TV

    THV

    THV

    
    By the way, notice that the voiceless characters are used where they are available, ie. in the first three sets of the 'D' row above. Oh, and, incidentally, the English 'TH' examples above demonstrate just how much more accurate the Cherokee syllabary is than English once you have access to these diacritics. Something similar could be done for English, I suppose, but I suspect it would be a bit more difficult what with all the complications of English spelling conventions.
  
  

  ---

  
  Here's another way these diacritic markers might be useful: I've noticed that different Cherokee dialects pronounce the QU-, DL-/TL-, and TS- rows differently. The nice thing about a system of diacritic symbols like this one is that variant pronunciations would be less able to break the standardized orthography into separate versions which might eventually become separate languages like French, Spanish, Italian, Portuguese, and Romanian did despite the long tradition of writing in Latin. Here are a few examples of what I mean:
  
  
  • QU- row: This consonant alternates between QU, GW, and glottal-stop + W. Take the word (A-QUA-DU-LI), for instance. Is is pronounced as AGWADULI, AGHWADULI, and even A'WADULI. Of course, the last version pretty much requires a re-spelling to (A-WA-DU-LI). Using the diacritic markings, however, we can capture most of these pronunciations with the same spelling. (Below, I will discuss how the glottal-stop variant could be incorporated into this spelling, as well). This would make it possible for an author to write in his or her own dialect or to accurately transcribe speech from a certain dialect or even from non-native speech. Here's a chart showing two of the three pronunciations described above plus two others that might possibly occur in speech by non-native speakers of, say, German or Arab extraction:
    
    

    A-QUA-DU-LI = 'I WANT':

    A-GWA-DU-LI = 'I WANT':

    
    

    A-KHWA-DU-LI = 'I WANT':

    A-GHWA-DU-LI = 'I WANT':

    
    
  • TS- row: I mentioned earlier that 'TS' is often pronounced as 'CH' in Cherokee. In the following example, though, it really means 'TS'. Sorry for the confusion, but the 'TS' row covers the sounds 'J' as in "Jones" or Badger, 'CH' as in "Cherokee" or Hatch, 'TS' as in "tsetse fly" or "cats", and 'DZ' as in "gadzooks" or "kids." The pronunciation varies due to a number of factors. From what I understand, the pronunciation of the word (TSA-LA-GI) itself varies between all these pronunciations depending on the dialect and, perhaps, other factors. Here's how the consonantal-alteration diacritic marks could represent each of these pronunciations:
    
    

    CHA-LA-GI = 'CHEROKEE':						TSA-LA-GI = 'CHEROKEE':

    
    

    JA-LA-GI = 'CHEROKEE':						DZA-LA-GI = 'CHEROKEE':

    
    
    
  • DL/TL- row: It is evidently common to pronounce TL as HL in contemporary Oklahoma Cherokee speech. The word for 'no', for example is spelled (TLA) but pronounced 'HLA'. I would like to suggest that the intrusive-H symbol I proposed in section four above have a slightly modified meaning when used with characters from the TL row. Here, instead of indicating an added 'H', it would signify that the 'T' is changed to 'H'. This provides some flexibility in that there are now two possible ways to spell the syllable HLA using the intrusive-H marker:
    
    

    TLA as HLA:			LA as HLA:

    
    
    I would expect that words that have traditionally been spelled using the syllable (TLA) could continue to be spelled that way and that words traditionally spelled with (LA) could likewise keep the same spelling. That way, those who pronounce TLA as TLA can still read texts written by those who pronounce it HLA and vice versa. Not explicitly marking the syllable with the intrusive-H marker would leave the reader the option of using either of the two possible pronunciations.
    
    The variant pronunciation A'WADULI (I want) could easily be reconciled with the spelling (AQUADULI) using a similar approach. Where the forward-positionedd glottal-stop symbol occurs with a character from the QU- row, this will mean that QU- is sounded to glottal-stop + W. In other words, the stop portion of the consonant would become glottalized in a way that is similar to how most Americans pronounce the word 'button' where a catch in the throat replaces the sound of 'T'. Here's what that would look like:
    
    

    A-'WA-DU-LI = 'I WANT':

  
  
  

  ---

  Examples:

  I've put together a short list of English names as they are commonly transliterated into Cherokee script along with their slightly divergent pronunciations in order to demonstrate the zero-vowel symbol I proposed above. With the addition of the zero-vowel diacritic, the pronunciations become completely phonetic:
  
  

  English:	Pronunciation:		Cherokee Script:

  ALEXANDER	ELIGSENDA

  CONSTANCE	KANSTANSI

  DARRYL	DEWAL

  GLEN	GILEN

  
  Now, if you wanted to more precisely specify certain other sounds without resorting to Roman script, you could utilize additional symbols, as well. For example, the Cherokee for DAVID is DEWID and is written as follows:
  
  

  DAVID	DEWID

  
  If you wished to notate the 'V' sound as it occurs in English, you could use the altered voiced/unaspirated consonant symbol in conjunction with the (WI) syllable as follows:
  
  

  DAVID	DEVID

  
  
  

  ---

  English Diacritics:

  Let me conclude by demonstrating with a simple example how such a system could actually be applied to English, as well, in order to render the pronunciation of certain names more legible. The digraph TH is used to represent two distinct phonemes in English. The words 'THY' and 'THIGH', both spelled with 'TH', differ in terms of the pronunciation of the 'TH' only, the former being voiced and the latter voiceless. Otherwise, the words sound alike. This is what linguists call a minimal pair. Linguists look for minimal pairs in order to determine phonemic boundaries in a given language. Here's another example: The words 'bus' and 'buzz' differ only in terms of the 's' and the 'z' sound. Therefore, 's' and 'z' constitute different phonemes in English. (Well, in truth, it's a little more complicated than that. Because in certain instances, 's' and 'z' are actually allophones of the same phoneme. For example, in the plural marker which can be pronounced /s/ (cats), /z/ (dogs), or /^z/ (foxes) depending on the context. But don't worry about that now).
  
  It is fairly difficult to predict which value a given instance of TH signifies, particularly in names. The TH in HEATHER, for example, is pronounced voiced while that of GAITHERSBURG (a city in the Maryland suburbs outside Washington, DC) is pronounced voiceless. Using the Cherokee symbols for voicing/aspiration, we can make it clear which phoneme is meant:
  
  
  
  
  With the addition of similar diacritic marks to indicate vowel quality, vocalic R, L, and N, silent E and H, accent, and a few others, we could render any English name or word completely phonetic. Here are a couple of examples:
  
  
  
  
  Using digits, we could even show primary and secondary stress as in the following example:
  
  
  You'll note that only ambiguous items have been marked above. There was no need to mark, say, the 'P' in 'possession' because there was really no alternative pronunciation possible in that instance. Again, I don't envision such a system being used in print so much as penned in by readers themselves in order to aid them in pronunciation or memorization of new vocabulary items or names.
  
  At this point, you may be saying to yourself that a simple and straightforward spelling reform would be a better approach to this problem in the case of English. In fact, I did develop such a system several years ago. And though I developed it specifically for use in English, it works equally well for Arabic, German, French, and several other languages I like to play around with. I would demonstrate it here, but in order to do so, I would need to create a special set of .GIFs as I did for the syllabary because the system uses a couple of diacritics that can't easily be affixed to normal text since I have no way of predicting what size a font a given visitor is using. There are some examples, though, that don't require diacritics, so I'll post a few of those below:
  
  
  • 'pan', 'pen', and 'pin' are all spelled as in normal English orthography as are 'sand', 'send', 'sent', 'sin', 'bend', 'bin', 'tin', and a whole host of other mono-syllabic words and even a few multi-syllablic words like 'random', 'seldom', and 'pistol'.
    
  • 'son' is the spelling used for both 'SON' and 'SUN'
    
  • 'som' is the spelling used for 'SOME'
    
  • 'ai' and 'main' are used to spell 'I' and 'MINE'.
    
  • 'kud' and 'wud' are used to spell 'COULD' and 'WOULD'.
    
  • 'hiz' and 'horz' are used to spell 'HIS' and 'HERS'.
    
  • 'senotor', 'kanobol', 'terobol', and 'ferwel' are used to spell 'SENATOR', 'CANNIBAL', 'TERRIBLE', and 'FAREWELL'.
  
  When I get more time, perhaps I'll post another paper detailing the whole system with examples of each phoneme, etc. If this is something that interests you, please e-mail me and I'll be happy to send you further information.
  
  

  ---

  Footnotes:

  ¹Incidentally, some linguists would argue that only phonemic distinctions need be made using diacritic marks. Yet the nasality marker I have proposed in section three and many of the others I have proposed elsewhere throughout the paper notate phonetic distinctions only. Having studied Cherokee as a second language, I feel it is useful to notate some phonetic values for learning purposes. A native speaker automatically adds nasality in the appropriate places. But a non-native student does not know which syllables to nasalize without some guidance. While such a sign would have no purpose in ordinary Cherokee-language texts, I see it as a potentially useful addition to language textbooks and dictionaries as an aid to correct pronunciation. Click here to return to where you were when you clicked to get here.
  
  ²In her comments to me, Dr. Bender suggested that short vowels be signified with no marking while only long vowels would receive an explicit marking. Coming from the Arabic tradition, though, I would prefer to have an explicit marker for every phenomenon so that the absence of a marker will always mean the item is ambiguous. In this way, every Cherokee document in existence today already conforms to this system whether explicit markers have been added to certain syllables or not. Click here to return to where you were when you clicked to get here.
  
  

  ---

  Proposed System of Auxiliary (Diacritic) Markers to Aid Children and Adult Learners in Mastering Cherokee Script / v1.01 / Sept 24th, 1999 / David Harris

  Visit these other Cherokee-language links:
  Cherokees of California: Cherokee Language Lessons
  The Cherokee Observer's Language Page (Observer font needed to read Cherokee portions)
  
  To access the home page of Dave's Bargain Basement, click here.