Annotating the texts: making explicit McQuown’s implicit analysis.

I present the texts here in the following manner, illustrated using line 1 of the text “Lixtu”:

(1) ʔantaá kšaqstiːpúːn liːštú_, šwií ⁿλankačučút_.
«{<ʔanta>≡ˑ} {<kšaqstiːpuːn>} {<liːštuh>},» «{<šwiː≡ˑ>} {ⁿ≡<λankačučut>}.»
ʔantaʔ nak=š–ʔaqstiːpuːn liːštuh š–wiː λankačučut
there LOC=3POSS–crest liːštú PST–sit lake
‘There, at the crest of Lištu, there was a lake.’

Line 1 is McQuown’s representation, with one addition: to make the prosodic system easier to follow for the reader, I have added to McQuown’s transcription the epenthetic nasals which are the result of a juncture process at the initial border of an Accentual Phrase when it is in medial position within an Intonational Phrase. McQuown considered prenasalized obstruents as allophones of the obstruents (1990:37–39) and, therefore, not to be transcribed in a phonemic representation. I write them as prenasalizations, so instead of leaving tlankačučút, as in the original, I write it as ⁿtlankačučút.

Line 2 is an explicit representation of the prosodic phenomena implicit in Line 1.

Line 3 gives morphemic boundaries of the underlying morphosyntactic representation.

Line 4 is a morpheme by morpheme annotation. For each unit, I give a free English translation, which is mine since the collection has only free translations in Spanish.

The symbols used are:

Line 1: McQuown’s representation.
“  _,” End of Intonational Phrase, continuing intonation.
“  _.” End of Intonational Phrase, final intonation.
Line 2:  Prosodic representation.
«…» Intonational Phrase (IP)
{…} Accentual Phrase (AP)
<…> Lex
H lengthening and devoicing of previous resonant
≡ˑ Prosodic vowel-lengthening
ⁿ≡ Prosodic epenthesis of nasal
IP, Continuing Intonation
IP, Phrase Final Intonation
Lexical Stress occurs on the last vowel of “…>”, unless it is written on the penult syllable.

Three levels are explicitly shown: the level of the Intonational Phrase (roughly equivalent to a clause); the level of the Accentual Phrase (consisting of one or several Lexes); and the level of the Lex (equivalent to a morphosyntactic word-form).

The three levels are stacking levels––that is, the Lexes are stacked within Accentual Phrases, which in turn are stacked within Intonational Phrases and exhibit Proper Bracketing (for example, the edge of an AP cannot be situated in the middle of a Lex). For CTot, the general rule is that the rules of the outermost layer apply. Whenever a higher and lower prosodic domain coincide, but require different phonotactics, the requirements of the higher domain always win out. This is a language-specific characteristic and is not at all the norm cross-linguistically (Ryan Bennet p.c.).

Translated into phonological rules, the rules of the IP level are domain-limit rules of the form shown in (7a), and the rules of the other two levels are domain-juncture rules, of the form shown in (7b). In these rules, A and B are segments, and B may be null. X and Y are strings of segments. Di and Dj stand for prosodic domains (see Selkirk 1980: 111–112).

(2) a. Domain limit
i) A→ B /[…X__]Di
ii) A→ B /[__X…]Di
b. Domain juncture
i) A → B / […[…X__]Dj [Y…]Dj …]Di
ii) A→ B/ […[…X]Dj [__Y…]Dj …]Di

So, showing the first unit of (1) above, observe that in the second line there are two IPs, the first one ending in the Lex liːštuh, the second one ending in the Lex λankačučut, both enclosed in angle brackets. Note also that those Lexes are simultaneously the only elements of their APs (and therefore they are enclosed in curly brackets), and that they are the last elements of their respective IPs—and, therefore, they have the closing angle quotes preceded either by a comma or by a period, indicating continuing intonation and final intonation, respectively. The end of an IP, then, is also the end of the two other prosodic levels, which means a) that words uttered in isolation are IPs and, as I said above, b) that the edge phenomena that prevail in CTot are those of the outermost layer.

As for stacking, observe that the first IP “«…»” is composed of three APs (enclosed in curly brackets) and the second IP is composed of two APs. In (3), we see that that APs can, in turn, consist of several Lexes, the elements enclosed in angle brackets.

(3) _maːskinaːniːlakwanhlak¢amaxanhšwanquːníːt_
maːski naː niː lakwan lak–¢amaxan š–wan–quː–niːt
although also NEG best PL–girls PST–be–PL.PRTPF
‘although also they were not the best girls (of their town)’ (Mc 8.42)
Illustrating the system.

To give the reader the gist of how the system works, I will show how the end boundary of a unit fluctuates according to the prosodic domain in which it is found. In the following examples watch the Lex čiškuʔ ‘man’ in three different prosodic domains.

At the end of an IP, and only in that domain, it retains its lexeme-final /ʔ/. A clear phonetic glottal stop can be heard in (10).¹

(4) with phonetic glottal stop at the end of IP
šlamaá ⁿčaːtám ᶮčiškúʔ_,
«{<šlama>≡ˑ} {ⁿ≡<čaːtam>} {ᶮ≡<čiškuʔ>},»
š–lamaː čaː–tam čiškuʔ
PST–live CL:HUMAN–one man
‘There was a man,’ (Mc 4.1.)

The glottal stop is dropped in the other two domains, so /čiškuʔ/ → [čišku], as can be seen (and heard) in (5) when Lex-final, AP-internal.

(5) čiškuʔ ‘man’ in Lex-final, AP-medial position:
papiːʔakítiː ᶮčiškušakwaá laː wíš_,
«{<papiː><ʔakit>≡iː} {ᶮ≡<čišku><šakwa>≡ˑ} {<laː>≡ˑ} {<wiš>},»
papiː ʔakit čiškuʔ ša–k–wa laː wiš
whether I man PST–be:PFV as you
‘If I were a man like you’ (Mc 4.9)

As we can see in the first line of (5), čiškušakwaá is written without intervening blanks and with one stress diacritic in Line 1, therefore it is one AP and, as such, it is enclosed by curly brackets in Line 2. However, it contains two Lexes, čiškuʔ and šakwa, which are enclosed in angular brackets in Line 2. So čiškuʔ is Lex-final, AP-medial and as we can see, in this domain its final glottal stop does not appear in the surface, it is dropped.

In (6) čiškuʔ ‘man’ appears in AP-final, IP-medial, and in this context the glottal stop is dropped too and the final vowel is then prosodicaly lengthened, so /čiškuʔ/ → [čišku] → [čiškuːː].

(6) čiškuʔ ‘man’ in AP-final, IP-medial position’
šakšiɬaá ᶮčiškuú ᵐpiːniːkskúxa_,
«{<šakšiɬa>≡ˑ} {ᶮ≡<čišku>≡ˑ} {ᵐ≡<piː><niː><kskúx[ḁ]>},»
š–ʔakšiɬ–yaː čiškuʔ piː niː šak–skúx–yaː laqačiniːn
PST–see–IPFV man that NEG PST–work–IPFV sometimes
‘He used to see the man, that he didn’t work’ (Mc 4.5)

In (7) I have edited the audio so that the reader can listen to the three realizations of čiškuʔ, IP-final in (7a), Lex-final, AP-medial in (7b), and AP-final, IP-medial in (7c).

(7) a. {ᶮ≡<čiškuʔ>},»              
b. {ᶮ≡<čišku><…
c. {ᶮ≡<čišku>≡ˑ}

These three examples show the unit-final processes.  All of them have prenasalization because the three happen to be IP-medial, AP-Initial.

In the next section I give the summary of prosodic processes represented in Line 2.

  1. And it is lexically conditioned—i.e., there are words that end in underlying glottal stops that surface in IP-final position.