A Historical Grammar of the Maya Language of Yucatan (1557-2000)
Page 8
consonants, it, too, yields consonant clusters when followed by roots. Once again, the phonological pro-
cesses for resolving such clusters are different for laryngeal and non-laryngeal initial roots. If the root- initial
consonant is not a laryngeal, then an epenthetic vowel [a] is inserted between the [k] and the following
consonant:
(16)
Input
Output
Gloss
c-botic [k-b’óʔotik]
ca-botik [ka-b’óʔotik]
we pay it
c-cuxtal [k-kuštal]
ca-cuxtal [ka-kuštal]
our lives
c-mehenob [k-mèehenóʔob’]
ca-mehenob [ka-mèehenóʔob’] our sons
c-naa [k-naʔ]
ca-naa [ka-naʔ]
our mother
c-patan [k-patan]
ca-patan [ka-patan]
our tribute
c-pixan [k-pìišan]
ca-pixan [ka-pìišan]
our souls
c-ɮolic [k-ȼolik]
ca-ɮolic [ka-ȼolik]
we explain it
c-yum [k-yùum]
ca-yum [ka-yùum]
our father
Modern Yucatec does not use [a]-epenthesis for separating the first-person plural clitic pronoun from
the consonant that follows it. Rather, it inserts a schwa [ə] between the [k] and the adjacent consonant.
Although [ə] has been attested as a contrastive vowel in the other Yucatecan languages (Lacandón, Itsaj,
and Mopán) and has been reconstructed as the sixth vowel in Proto-Yucatecan (Campbell 2000:345; Fisher
1973:110; Justeson 1986:6–7), it does not have phonemic status in Modern Yucatec, and there is only indi-
rect evidence for its existence in Colonial Yucatec (see 2.3.3. below).
In the absence of a symbol for schwa in the Colonial orthography, the scribe could have selected a as
an alternative (the most common way of representing schwa in Colonial texts [see 2.3.3. below]). However,
a was not available for this purpose because it served as the epenthetic vowel [a] before non-laryngeal
initial roots (see [16] above]. Instead, the scribe represented the first-person plural clitic pronoun before
glottal-stop initial roots by c alone, as in the second column of (17), thereby maintaining an orthographic
distinction between the two forms of the pronoun:
(17)
Phonetic Input
Colonial Spelling
Aberrant Spelling
Gloss
k-ʔohel
c-ohel
k ohel [k’-ohel]
we know it
k-ʔilah
c-ilah
k ilah [k’-ilah]
we saw it
k-ʔáʔalik
c-alic
k alic [k’-áʔalik]
we say it
And although the initial glottal stop was never represented explicitly, it can be inferred from examples of c
becoming k (the velar ejective) in the same contexts (as shown in the third column of [17] above). The same
process occasionally occurs in Modern Yucatec, but the more common pattern is for a schwa to be inserted
between the prefix and the glottal stop (Orie and V. Bricker 2000:295).
28 PHONOLOGY
Another kind of consonant cluster is produced when pronominal c or ca are preceded by ti (phonetic
[tiʔ]). ti had two functions in Colonial Yucatec, representing the general preposition meaning ‘in, at, on, to,
from,’ as well as serving as the completive aspect clitic with perfective verb stems.
Although the glottal stop at the end of [tiʔ] was usually deleted when it was followed by a consonant-
initial prefix or root, this was only rarely the case when followed by the first-person plural clitic [ka]. Rather,
[tiʔ ka] normally became [tak]. Perhaps on analogy with what happened when [tiʔ] was followed by the VC
pronouns [aw] and [uy] (as well as the VCC pronoun [inw]), where the [i] in [tiʔ] assimilated to the vowel
in the pronoun (e.g., [tiʔ u] became [tu], and [tiʔ a] became [ta]) (see 2.3.1. below), both the vowel and the
glottal stop in [tiʔ] were deleted before [ka], and [a] metathesized with [k], yielding [tak]:
(18)
Input
Output
Gloss
ti ca-kamah [tiʔ ka-k’amax]
tac-kamah [tak-k’amax]
we received it
ti ca-lacal [tiʔ ka-láakal]
tac-lacal [tak-láakal]
all of us
ti ca-menel [tiʔ ka-mèenel]
tac-menel [tak-mèenel]
because of us
ti ca-pixan [tiʔ ka-pìišan]
tac-pixan [tak-pìišan]
in our souls
ti ca-ɔah [tiʔ ka-ȼ’áax]
tac-ɔah [tak-ȼ’áax]
we gave it
ti ca-xotah [tiʔ ka-šotax]
tac-xotah [tak-šotax]
we cut it
Thus, having reduced [tiʔ ka] to [tka], the strategy chosen was to metathesize the vowel with the second
consonant, rather than inserting an epenthetic vowel between the [t] and the [k]. For reasons explained in
Chapter 4, the metathesized form of the first person plural clitic dropped out of use during the second half
of the eighteenth century. However, it survives today in Lacandón as [ək], the form of the pronoun used
with non-laryngeal initial roots (Bruce 1968:48).
[tak] can also be attested before laryngeal-initial roots:
(19)
Input
Output
Gloss
ti c-alah [tiʔ k-ʔáʔalax]
tac-alah [tak-ʔáʔalax]
we said it
ti c-etsah [tiʔ k-ʔéʔetsax]
tac-etsah [tac-ʔéʔetsax]
we showed it
ti c-ilah [tiʔ k-ʔilax]
tac-ilah [tak-ʔilax]
we saw him
Because the prefix in these examples is [k] rather than [ka], the process is one of vowel epenthesis, not
metathesis.
2.2.2.2. SUFFIXES. The consonant clusters produced by suffixation are resolved by several phonological
processes: debuccalization, deletion, assimilation, metathesis, and epenthesis.
2.2.2.2.1. DEBUCCALIZATION. In Modern Yucatec, the ejectives [b’], [ȼ’], and [k’] are often debuccalized when they are followed by the causative suffix [s]:
(20)
Input
Output
Gloss
hóok’-s
hóʔo-s
remove
líik’-s
líʔi-s
raise
lúub’-s
lúʔu-s
fell
luk’-s
lúʔu-s
remove
púuȼ’-s
púʔu-s
cause to flee
PHONOLOGY
29
In these examples, [b’], [k’], and [ȼ’] have been reduced to [ʔ], yielding the consonant cluster [ʔs]. This clus-
ter is then eliminated by inserting a copy of the root vowel between [ʔ] and [s].
My Colonial sources contain examples of two causative transitive verbs that seem to behave in the
same way, one of which is listed in (20) above:
(21)
Input
Output
Gloss
hok-s [hóok’-s]
ho_-s [hóʔo-s]
remove
nab-s [nab’-s ?]
na_-s [náʔa-s ?]
anoint, varnish, smear
The ejectives that appear at the end of hok and nab may have been replaced by a glottal stop, which of
course was never represented in the Colonial orthography.
Debuccalization can also occur when the suffix rather than the final consonant is an ejective. In roots
ending in non-laryngeal consonants, the passive stem is formed by suffixing -b (= phonetic [b’]) to the root.
In Modern Yucatec, the resultant cluster is modified, first by metathesizing [b’] with the root-final conso-
nant, then by debuccalizing the ejective to [ʔ], and finally by separating the two consonants by inserting a
copy of the root vowel between them:
(22)
Input
Output
Gloss
b’on-b’
b’óʔon
be painted
ȼik-b’
ȼíʔik
be respected, obeyed
kuč-b’
kúʔuč
be carried
mol-b’
móʔol
be gathered
p’at-b’
p’áʔat
be left, abandoned
sop’-b’
sóʔop’
be sold cheaply
šel-b’
šéʔel
be opened
tak’-b’
táʔak’
be stuck
wuȼ’-b’
wúʔuȼ’
be folded, bent
The following examples suggest that the same processes were at work in Colonial Yucatec:
(23)
Input
Output
Gloss
kal-b [k’al-b’]
kaal [k’áʔal]
be closed, covered, imprisoned
mol-b [mol-b’]
mool [móʔol]
be gathered
pot-b [pot-b’]
poot [póʔot]
be perforated
ɮic-b [ȼik-b’]
ɮijc [ȼíʔik]
be respected, obeyed
uk-b [ʔuk’-b’]
uuk [ʔúʔuk’]
be drunk
If, on the other hand, the final consonant of the root was a laryngeal, then the suffix was not debuccalized.
Instead, the laryngeal was deleted, and the ejective remained unchanged:
(24)
Input
Output
Gloss
chii-b [čiʔ-b’]
chiib [číʔib’]
be bitten
choo-b [čoʔ-b’]
choob [čóʔob’]
be cleaned
cħaa-b [č’aʔ-b’]
cħaab [č’áʔab’]
be taken, seized
le-b [leh-b’]
leeb [léʔeb’]
be lassoed
30 PHONOLOGY
paa-b [paʔ-b’]
paab [páʔab’]
be broken
too-b [toʔ-b’]
toob [tóʔob’]
be wrapped
ɔa-b [ȼ’ah-b’]
ɔaab [ȼ’áʔab’]
be given
xe-b [šeh-b’]
xeeb [šéʔeb’]
be vomited
zi-b [sih-b’]
ziib [síʔib’]
be presented
The contrast between the treatment of laryngeal-final roots and roots ending in other consonants implies
that laryngeals are weaker than other consonants (see 2.3.1. below).
2.2.2.2.2. VOWEL INSERTION. The passive suffix can also be attached to derived transitive stems, which
already end in two consonants, the one terminating the root and the derivational suffix, either -s or -t. The
addition of -b produces a triconsonantal cluster. The solution here was to insert a vowel—always [a]—be-
tween the second and the third suffix:
(25)
Input
Output
Gloss
cims-b [kíims-b’]
cins-ab [kíins-ab’]
be killed
hoks-b [hóok’s-b’]
hoks-ab [hóok’s-ab’]
be removed
mans-b [máans-b’]
mans-ab [máans-ab’]
be transferred
ocs-b [ʔòoks-b’]
ocs-ab [ʔòoks-ab’]
be inserted
sats-b [sáʔats-b’]
sats-ab [sáʔats-ab’]
be lost
sihs-b [síihs-b’]
sihs-ab [síihs-ab’]
be engendered
tals-b [tàals-b’]
tals-ab [tàals-ab’]
be brought
(26)
Input
Output
Gloss
hant-b [hàant-b’]
hant-ab [hàant-ab’]
be eaten
kakt-b [k’áak’t-b’]
kakt-ab [k’áak’t-ab’]
be roasted
kult-b [k’ult-b’]
kult-ab [k’ult-ab’]
be worshipped
lacht-b [láʔačt-b’]
lacht-ab [láʔačt-ab’]
be scratched lightly
meyaht-b [meyaht-b’]
meyaht-ab [meyaht-ab’]
be served
mist-b [míist-b’]
mist-ab [míist-ab’]
be swept
ɔibt-b [ȼ’íib’t-b’]
ɔibt-ab [ȼ’íib’t-ab’]
be written
tumt-b [túumt-b’]
tumt-ab [túunt-ab’]
be tested
Other phonological processes, such as debuccalization (see 2.2.2.2.1. above) and liquid-deletion (see
2.2.2.2.3. below) could then be applied to modify the biconsonantal cluster that remained.
2.2.2.2.3. LIQUID DELETION. In root- and suffix-final position, [l] is frequently deleted when it is followed
by another consonant:
(27)
Input
Output
Gloss
colel-bil [kóʔolel-b’il]
cole_-bil [kóʔole_-b’il]
lady
hol can [xol-kan ?]
ho_ can [xo_ kan ?]
warrior
kul che [k’ul čeʔ]
ku_ che [k’u_ čeʔ]
cedar
pom-ol che [pom-ol čeʔ]
pom-o_ che [pom-o_ čeʔ]
Jatropha gaumeri Greenman
ho ppel hab [hóʔoh
ho ppe_ hab [hóʔoh
five years
p’éel háʔab’]
p’ée_ háʔab’]
PHONOLOGY 31
cil-mac ol [kíʔil-mak ʔóol]
ci_mac ol [kíʔi_-mak ʔóol]
joy, happiness
hol-na [hòol-nah]
ho_-na [hòo_-nah]
door
tal-s-ic [tàal-s-ik]
ta_-s-ic [tàa_-s-ik]
bring it
bel-t-ic [b’èel-t-ik]
be_-t-ic [b’èe_-t-ik]
do it
ol-ɮ-il [ʔóol-ȼ-il]
o_-ɮ-il [ʔóo_-ȼ-il]
poor
bal-x [b’áʔal-š]
ba_-x [b’áʔa_-š]
what?
This suggests that [l] at the end of roots and suffixes is weaker than other consonants, since it can be
deleted before virtually any other consonant, including stops (c, t), affricates (ch, ɮ), ejectives (b, pp), laryn-
geals (h), fricatives (s, x), and nasals (m, n).
The weakness of [l] in root- and suffix-final position is also evident in Modern Yucatec, even when [l]
is not followed by another consonant. For example, the expression miš b’áʔal ‘nothing’ is often realized as
miš b’áʔah in many dialects today, and [l]-final suffixes such as -tal become [h]-final in the eastern dialects
of the language: e.g., táan u saktal ‘it is becoming white’ — > táan u saktah. Furthermore, it can be seen
by looking at the history of [l] in the deictic enclitics, la (= phonetic [laʔ]) ‘proximal�
�� and lo (= phonetic [loʔ])
‘distal,’ that initial [l] also weakened through time. It began to disappear during the late eighteenth century
and is attested today only when it immediately follows an initial deictic ending in a laryngeal. In such cases,
the laryngeal may be deleted (as in [toʔ loʔ] — > [to_ loʔ] ‘[out] there’), or a copy of the vowel in the initial
deictic may be inserted between the laryngeal and [l] (e.g., [heʔ laʔ] — > [héʔe laʔ] ‘here it is’ (cf. 1.2.4.2. in Chapter 15). The deictic enclitics in Modern Yucatec are normally -aʔ and -oʔ.
2.2.2.2.4. INTERROGATIVES WITH -X. In Colonial Yucatec, certain nouns and particles can function as rela tive
pronouns, including bal ‘thing, what,’ bic ‘how,’ mac ‘person, who,’ and tub ‘where.’ To each of these words,
the interrogative x (= phonetic [š]) may be suffixed, converting them into interrogative pronouns. The
consonant clusters produced by this suffix are frequently simplified by deleting the root-final consonant:
(28)
Input
Output
Gloss
bahun-x [b’axun-š]
bahu_-x [b’ahu_-š]
how much?
bal-x [b’áʔal-š]
ba_-x [b’áʔa_-š]
what?
bic-x [b’ik-š]
bi_-x [b’i_-š]
how?
mac-x [máak-š]
ma_-x [máa_-š]
who?
tub-x [túʔub’-š]
tu_-x [túʔu_-š]
where?
The input and output forms both existed in Colonial Yucatec. All traces of the input forms have disappeared
in Modern Yucatec.
2.2.2.2.5. NASAL ASSIMILATION. In Modern Yucatec, roots with [m] as the final consonant become [n] be-
fore dental, alveolar, and palatal consonants and [ŋ] before velar and labiovelar consonants, whereas roots
with [n] as the final consonant become [m] before bilabial stops and ejectives. The same kind of assimila-
tion occurred in Colonial Yucatec:
(29)
Input
Output
Gloss
cim-s-ic [kíim-s-ik] cin-s-ic [kíin-s-ik]
kill him
tum-t-ic [túum-t-ik] tun-t-ic [túun-t-ik]
test it
num-ya [núum-yah] nun-ya [núun-yah] misery
cum-ku [kùum-k’uh] cun-ku [kùun-k’uh] kiln
32 PHONOLOGY
can-bal [kàan-b’al] cam-bal [kàam-b’al] learn
chun-p-ah-al [čun-p-ax-al] chum-p-ah-al [čum-p-ax-al] begin
hun-ppel [hun p’éel] hum-ppel [hum-p’éel]
one (thing)
The only direct evidence of [ŋ] as an allophone of [n] in the written sources occurs in an example of the
first-person singular clitic pronoun before a glottal-stop initial stem, where it precedes the labiovelar con-
sonant: yngu almahmail (phonetic iŋw áʔalmaxmayil]) ‘I have reported it.’