The sounds of Mochica are not known with certainty due to the differing transcriptions used by different authors. In addition, the two primary sources on Mochica were recorded over 200 years apart, with significant phonetic changes having occurred during that time.
Vowels Carrera Daza (1644) Carrera Daza distinguishes six vowels in Mochica, represented as , with the latter sometimes being written as in different editions of his grammar. This vowel is characterized as "begin[ning] as an
e and end[ing] as a
u, in such a way that there are two vowels in one". A diacritic is used to indicate length, though it is mainly restricted to a few roots ( 'big') and the ending , used in the
genitive construction.
Middendorf (1892) In contrast, Middendorf recognizes 17 vowels, including 11 plain vowels, two "impure" vowels, written as and , and four diphthongs. Middendorf admits that he was never able to correctly pronounce the impure vowels, but characterizes them similarly to the of Carrera Daza. is described with a very fleeting
u sound, such that in rapid speech it would sound similar to or . is described as reminiscent of the diphthong . commonly found in
Amazonian languages. Other hypotheses have given realizations of and .
Comparison A comparison table of the vowels as written in Carrera Daza and Middendorf is given below.
Table Thus, a vowel system of Mochica would resemble the table below:
Consonants The orthography used for Spanish at the time of the writing of Carrera's grammar was very chaotic and inconsistent. One may therefore never be quite sure on the nature of consonants, especially sibilants, in Spanish colonial texts of the time.
Sibilants The symbols used to represent Mochica
sibilants in Carrera (1644) are and . The grapheme is described unanimously among Mochica scholars as representing . All but two authors coincide in their description of the
phoneme represented by or . The only sources to disagree are Stark (1968), which interprets them both as different
allophones of , being and respectively. Hovdhaugen (2004) interprets them as having a palatal pronunciation, which is represented as . Carrera (1644) explains the distinction between and as "they have to be pronounced between them both, hurting on the last one as in ssonto, amoss." The use of "hurting" is derived from Nebrija's
Gramática de la lengua castellana, who described consonants and vowels using this definition: "they were called vowels because they have voice by themselves without mixing with other letters, the others were called consonants because they cannot sound without hurting the vowels". Thus, combining the two descriptions, it may be inferred that "sounded only when the second (last) would affect the contiguous vowel", representing a sound like or .
Digraphs and trigraphs Digraphs and
trigraphs used in Mochica are , , and . The digraph is derived from a normal but with the inverted to represent the difference. Middendorf (1892) describes his corresponding sound, written as , as similar to the German . Hovdhaugen (2004) interprets it as , Stark (1968) as , Torero (1997, 2002) as , Eloranta (2013) and Michael et al. (2015) as , Cerrón-Palomino (1995) as and Adelaar as . The trigraph is described as representing a sound difficult to pronounce by Carrera: "These [words] one pronounces starting with T, hurting on the Z and on the vowel, that comes after H, so that it does not say
cha but
tzha." He further describes the articulation of it as having the "tongue touch[ing] the palate next to the teeth". Middendorf describes it as similar to the German . The trigraph is, according to Carrera, articulated as "[t]he X preceding consonant has to be pronounced hurting between both in a soft way, attaching the tongue to the palate, in such a manner that the sound of the first letter, the vowel, may come out through one side and the other of the mouth". Stark (1968) identifies this sound as , Torero (2002) as a
voiceless post-palatal lateral fricative , Hovdhaugen (2004, 2005) as , Adelaar (2004) as , and Salas, Eloranta and Michael et al. (2015) as .
Other consonants The following consonants are unanimously agreed upon by all sources: • voiceless
plosives , , ; is added by Hovdhaugen to represent the digraph • voiced plosive , also interpreted as by Adelaar and Hovdhaugen, and by Stark •
nasals , , , , though is replaced by a pre-palatal nasal by Torero (2002) • the
rhotic , an alveolar
trill; a tap is also added by Stark and Cerrón-Palomino • a labial
fricative, either or • the palatal
glide • the
postalveolar fricative , although is also given by Stark as an alternative to represent the grapheme • the postalveolar affricate , which is replaced by Stark with • the
laterals and ; Torero adds and interprets as pre-palatal. shifted to around the 18th and 19th centuries.
Table Michael et al. (2015) summarize their interpretations of Mochica in the following table.
Comparison A comparison of various interpretations for the letters is given below. == Typology ==