Having taken a look at each of the three main orthographic traditions, we can now compare them and propose a new orthography that will incorporate the strengths of the three traditions and provide alternatives to what are perceived in Chapter 2 as their weaknesses.  Each of the following aspects will be considered in turn: consonants, simple oral and nasal vowels, compound vowels and semi-vowels, nasalization, and stress.


The following chart is a comparison of the three traditions in terms of the consonants they consider as existing in the language, and the symbols they use to represent those consonants.  Also included are the symbols proposed for the new orthography (except those choices that require some explanation).

b    b   b     b
d    d   d     d
f    f   f     f
g    g     g     g
h    j   h
k    c, k, qu    k
l    l     l     l
m    m     m     m
n    n    n     n
p    p     p     p
r     r     r     r
s    s     s     s
t    t     t     t
–    w     w
–    y     y
c    ch, c   ch
j (in loan words)
–    ? (nasal semi vowel) ?

There is agreement on eleven of these consonants: b, d, g, f, l, m, n, p, r, s, and t.  These would therefore remain unchanged in the new orthography.  This leaves a total of seven other possible consonants to which we will now turn our attention.  In the case of three of these the disagreement lies not in the existence of the phonemes in question,but in the choice of symbols to represent them.  The other four have to do with differences in the phonological assumptions that underly the orthographies.


h     j     h     h, j
k’    c, k, qu    k     k
c    c     ch     ch
_    w     w     w
_    y     y     y
–    –     j (in loan words)  –
–    ? (nasal semi vowel) ?   –

It is obvious that it is because of social reasons, specifically the influence of Spanish orthography, and the goal of acceptability to authorities in Guatemala, where the SIL studies of Garifuna were done, and Honduras, where the majority of Garifuna people live, that the SIL chose j as the symbol for the phoneme that Taylor and S.J.C. write as  h.   This must also be the motivation for the use of c in most instances, and qu in a few, instead of k which is used only in English loan or foreign words.  But even though the importance of the social goals of orthographies cannot be ignored or downplayed, the aim of universality for the spelling system being proposed here requires that the larger social context, specifically the Garifuna social context, be considered.  h does exist in the Spanish alphabet, although it is usually ‘silent’, so the symbol would not be new to persons literate in Spanish.  What they would have to do would be to become accustomed to assigning it the value that they usually give to j when reading Spanish.  On the other hand, if we chose to retain j it would pose more of a problem for persons literate in English, but not Spanish, since they associate that symbol with an entirely different sound.  It is
therefore proposed that h be used for the phoneme in question.  However, this may turn out to be the most controversial grapheme and we may end up agreeing to use j in the sphere of Spanish influence and h elsewhere.

The choice between c, k, are borrowed by SIL from Spanish, but the fact that it does use k as well, albeit only in loanwords, indicates that it is perceived as a proper representation of the phoneme even within the spere of Spanish influence.  It is proposed that k be used to represent all occurrences of the consonant and that c and qu be dropped.
ch is bound to be more acceptable than c for the consonant that Taylor describes as “varies freely between the sound of ‘ch’ in ‘church’ and ‘sh’ in ‘shut'”.  ch is used in both Spanish and English writing.

It was pointed out in our discussion of S.J.C. orthography that [dz] does not exist as a phoneme in Garifuna.  For this reason, and the possibility that j may be retained as an alternative symbol to h in Spanish speaking countries, j as an equivalent of [dz] will not be listed in our proposed alphabet.  Foreign words, as opposed
to loanwords, can be written exactly as they would be written in the source language.

w and y will be considered below as semi-vowels, and ? will be given some attention when we look at nasalization.  However, we can conclude on the basis of the current practice of the SIL and S.J.C. traditions, and in the interest of orthographic elegance, that w, y, and ? will have to be retained whether they are phonemes, or simply variants of /u/, /i/, and /i/ as Taylor with some justification claims.

Simple Vowels

i    i    i     i
e   e    e     e
a    a    a     a
–    o    o     o
u   u    u     u
o   ?    ?     ?

Both the SIL and the S.J.C. traditions posit a six-vowel system, and they use the same symbols.  The only difference is that S.J.C. incorrectly ascribes to them the values associated with them in English, while SIL, as does Taylor, equates them with the vowels in Spanish.  Taylor denies the existence of a phonemic /o/ in Garifuna and uses that symbol for the phoneme that SIL and S.J.C. write as ?.  It was demonstrated in chapter two that /  / is a phoneme in Garifuna and that the symbol o is needed to represent it in such words as kopu (cup), fonu (thing), to (this), and ino (no).  ? is
a more practical choice for the unrounded mid-back vowel, since unlike o it exists neither as a grapheme nor as a phoneme in English or Spanish.

Taylor, SIL, and S.J.C. all agree that the simple vowels all have nasalized counterparts.  Therefore, six simple nasal vowels will be recognised and will normally be represented as in, en, an, on, un, and ?n.

Compound and Semi-Vowels

ie  ie (ye) ie (ye)  ie (ye)
ia  ia (ya) ia (ya)  ia (ya)
io     (y?)    (y?)     (y?)
iu  iu (yu) iu (yu)  iu (yu)

uo     (w?)    (w?)     (w?)
ua  ua (wa) ua (wa)  ua (wa)
ui  ui (wi) ui (wi)  ui (wi)
ue  ue (we ue (we)  ue (we)

ei  ei  ei   ei
ai  ai  ai   ei

au   ou  au   ou
au  au  au   au
ao  a?  a?   a?

There is general agreement about the first eight compound vowels.  The major difference lies in the fact that both SIL and S.J.C. re-write some of Taylor’s i’s and u’s as y’s and w’s respectively.  Taylor’s position that there is no opposition between i and y or between u and w, and that one is simply a variant of the other appears to be correct.  His refusal to use y and w are therefore supported by the general phonemic goal that “There should be a one to one correspondence between each phoneme and the symbolization of that phoneme”  (Pike 1968, pg.208).  Yet, it must not be forgotten that the purpose of that or any other orthographic principle that may be adopted is to make the language easy to write and read, and to make the orthography acceptable to the people who are expected to adopt it or encourage its use.  I contend that there are too many unbroken vowel sequences in Taylor’s orthography, and that such words as ueiu, huia, iaiaua, uaua, and uaibaiaua are easier to read if written thus:  weyu, yeyawa, wanwa, and waibayawa.

Whether or not the founders of the SIL and S.J.C. orthographic traditions are aware of it, their conversion of Taylor’s /i/ and /u/ to y and w respectively is not arbitrary but is rule governed.  The rules can be stated as follows:

1 If /i/ or /u/ and the following vowel are mono-syllabic and occur in word initial position, the /i/ and /u/ are written as y and w respectively.

2 When /i/ or /u/ occur between two other vowels they are written as y and w respectively.

These rules make it possible to re-write many of the words from Taylor’s script so that they appear to have more consonants, thereby making them easier to read.  They also make it possible to account for the difference in the way the following words are written:
war?? (our child)
u?ra? (madness)
?ara? (togetherness)
wara (our heat)

Taylor, SIL, and S.J.C. all concur on the existence of /ei/ and /ai/ as two separate compound phonemes and all represent them in the same way.  Given this agreement, it is unfortunate that I am forced to disagree with their analysis and argue that there is no opposition between the two, and that all instances of [ei] and [ai] should be written as ei.  It is difficult to explain how the identity of [ei] and [ai] could have escaped the followers of the three traditions.
Taylor and S.J.C. use au where SIL employs ou.  The SIL choice is clearly preferable for two reasons.  Firstly, ou is phonetically more faithful to the sound in question and, secondly, SIL, unlike the other two, is able to capture a necessary distinction between /ou/ and another compound vowel which can be better represented as /au/; e.g. touba (her side) and tauba (it will be with her).

Finally, in the choice between ao and a?, the latter has to remain in our orthography.  It will be recalled that ? was considered preferable to Taylor’s ‘o’ as a symbol for the vowel in question.


As we have seen, Taylor tries to deal with nasalization by indicating nasality on those vowels, by means of an inferior hook.  This practice has two main disadvantages; firstly, it is not so easy to keep track of individual nasal vowels, a task which is not made any easier by the fact that nasality may also be environmentally conditioned.  Secondly, there is something distasteful, even unnatural, about an orthography that has hooks sticking from underneath many of the vowels, a practice which the would-be users do not see associated with other writing systems that they are familiar with.
The SIL and S.J.C. traditions use a method that is more acceptable. Here a nasal consonant n (and also m in the case of S.J.C.) is used to indicate nasality of all simple vowels and most vowel sequences.  The n (or m) is placed after the vowel or vowels, which is to say that it goes at the end of the syllable, and since the language tolerates very few consonant clusters, the n which is followed by a consonant can only be a signal for nasalizing the preceding vowel or vowels.  Unfortunately, this neat arrangement is slightly complicated by the use of ? as a consonant by S.J.C., and as a
nasalized glide, y, by SIL, and also to indicate nasality in some vowel sequences, including ie, ia, iu, and i?, which are written i?a, i?e, i?u, and i?u.  The problem here is that the use of ? in these vowel sequences is unnecessary since they can just as easily be written as ien, ian, iun, and i?n.  It is proposed that the use of ? be restricted to those instances when it indicates a nasalized [y] or where the use of an n to indicate nasality would require the use of y as well.

For example, yei (there), ay?rein (there are), gay?giru (she’s still a virgin), lig?ar?g?yei (he’s the only one) can be written thus:

?ei    yein
a?arein or  ayanrein
ga?ugiru   gay?ngiru
lig?ar?gu?ei  ligiar?g?yein

I believe that the second solution which uses n to indicate nasalization more faithfully captures the actual pronunciation of the words.  In reading, the nasalizing effect of the grapheme n goes backward, so that the preceding vowels and, to a lesser extent, the preceding semi-vowel, are nasalized.  On the other hand, ? tends to be perceived as a hard consonant so that even when it is used as an equivalent of y so much of the nasality seems to be concentrated in it that it is not carried over sufficiently into the succeeding vowels which should be the true bearers of the nasal quality.  Indeed it can be argued, and this is my position, that what the SIL tradition perceives as a nasal glide, y, and the S.J.C. tradition seems to view as a palatal nasal consonant, is actually an oral glide [y] which appears to be nasal because of the influence of the nasal vowels that follow.

But even though I contend that the second method which employs both y and n instead of ? is a more faithful representation, it is proposed that ? be retained for two reasons.  First, ? is so entrenched in Garifuna writing that there is bound to be much resistance to its elimination.  It is perceived by the average person as a consonant.  This makes it more visible than say ai (which it is argued should be dropped) and easier to deal with than a vowel or a semi-vowel with an elusive nasal topping!  Secondly, ? is only one symbol while the second method would require the use of two symbols.  The additional symbol can make a disgusting difference when we are dealing with long words like lig?ar?g?yein/lig?ar?g??ei (he’s the only one), or layoundaguar?g?yan/la?oudaguar?g??a (he’s only kneading).


Of the three traditions Taylor gives the best description of stress in Garifuna.  Basically he makes two statements that can be regarded as general rules for accentuation.

1 In words with two syllables, the stress tends to fall on the first syllable.

2 In words with three or more syllables, stress tends to fall on the second syllable and thereafter on every third syllable.

This means that the accent need be written in only in the case of the exception to these rules.