By Deanna Jager
Traditionally, one way of dividing languages is into either tonal or stress languages. This division refers to the way in which pitch is implemented. In stress languages, like English and Spanish, pitch is one correlate of stress, which is used to mark the prominence of syllables in words. In tone languages pitch is used to distinguish meaning among words.
This means that in English the word ma can be realized with a variety of pitches such as:
1) high má
low mà
falling-rising mǎ
And all versions will still mean ‘mother’. On the other hand, in Mandarin, a tonal language, these types of pitch changes result in lexical differences:
2) high má ‘hemp’
low mà ‘scold’
falling-rising mǎ ‘horse’
Tonal languages are further subdivided into tone and pitch accent. Tone languages, like Mandarin and Zapotec, mark each syllable with a specific tone (pitch) and changing the tone can change the meaning of a word, as illustrated above. Pitch accent languages, like Japanese and Swedish, mark each word with a specific tonal melody and changing the melody can change the meaning of the word.
Blackfoot is a pitch accent language and I will be looking at the function and phonology of pitch. As the tonal melody can correlate to many features, including length, pitch height and loudness, altering any of these features has the potential to indicate a changed melody. Van Der Mark (2001) analysed the acoustic phonetics of prominence in Blackfoot as a combination of average amplitude (loudness) and F0 peak (pitch height), but remarks that amplitude is an unexpected correlate that cannot be fully explained. I will be looking solely at relative pitch height as indicating prominence/pitch accent.
Relativity is an important concept in looking at pitch accent. Pitch is said to be either high or low, but these are not absolute terms that correspond to a specific measurement of Hertz. A high pitch is high in relation to the surrounding lower pitches, as a low pitch is low in relation to the surrounding higher pitches. As all the data elicited for this project was from a female speaker, the pitch range will be relatively higher than that of a male speaker, but the salience of pitch will convey the same meaning for all speakers.
The tonal melody of the word is realized as a relative high tone in Blackfoot and can occur on any syllable, as the following list illustrates.
3) first syllable ómahk big
ááattsistaa rabbit
píítaa eagle
4) second syllable akáwu many
kokóto ice
i’táámpsi happy
5) third syllable i’naksípokaa little baby
ksiistsikó day
6) fourth syllable asikkiááyu black bear
Since Blackfoot syllable structure allows for long nuclei, the language has long vowels and diphthongs, which can carry pitch on either or both mora, as the following list illustrates.
7) first mora long vowel áa yes
pokáa child
pináap east
diphthong ponokái elk*
awáohkaani game*
niisói four*
8) second mora long vowel aoówahsin food
poós cat*
9) both morae long vowel míín berry
i’táám happy
óópaihp waist
diphthong kitáísiksikimi your tea
iihtáóhpommao’p money
iitáókakihtsimaan courthouse
Pitch accent is lexical, meaning that it contributes to a word’s meaning. As the following list of minimal pairs illustrates, if you change the location of the high pitch (as in (10)) or change the high pitch altogether (as in (11)) you end up with a different lexical entry.
10) ápssi arrow
appsí white buffalo berry
11) taksuwuu I’m going on a war party
taksuwúú I’m going into the water
12) saaámi head dress
saaámi medicine
note: although these both carry pitch on the same vowel, the pitch on head dress is significantly higher.
13) niitsáápiikoan French person+
niitsaapiikoan Frenchmen+
Although pitch can be lexical, it can shift its position from one syllable to another in various situations without changing the meaning. To analyse the phonology of pitch I will look at three situations: compound nouns, affixing in verb paradigms and other prefixes.
Compound Nouns
14) kipitáaakii old woman kipita old + aakíí woman
15) i’naksípokaa small child i’nak small + pokáa child
16) náápiohkiin wine náápi trickster + aohkíí water
17) innóóhsoyis spoon+ inno long + mohsoyís tail
As these compounds illustrate, pitch that occurs on the second noun in a compound shifts left to the morpheme boundary, attaching to the leftmost syllable in the leftmost morpheme, when combined with these nouns. For example, (14) shows that the lexical pitch of aakíí woman shifts to the final syllable of the affixed word kipita old, which was previously accent-less.
This is not unexpected, as prominence (in this case pitch) often marks boundaries. But why do we see these specific shifts? According to Kaneko (1994), who looked at pitch in Blackfoot nouns, pitch accent in Blackfoot is weight sensitive; this means that pitch is attracted to heavy syllables. Kaneko argues that both long nuclei (long vowels and diphthongs) and closed syllables are heavy (ie: coda consonants are moraic).
When words are made up of only light syllables, the default accent assignment is on the final syllable. This is also shown in (15), assuming the syllabification is: /i’.na.ksí.po.kaa/ (see Kaneko: 1994 for explanation of /ks/) . When there is a heavy syllable, as in (10) this attracts pitch; this is why we don’t see */naa.pí.oh.kiin/. Example (17) is the same; we don’t see */ ín.nooh.so.yis/ because the heavy syllable attracts the pitch.
Verb Paradigms
In Blackfoot, the durative aspect is marked with an ‘a-’ prefix. As illustrated below, this marker attracts pitch. For example, in (19) the pitch on okát sleep is on the final syllable, but when the durative marker is added it carries pitch: nitáiooka I am sleeping.
18) nitákipoosk I dance
nitsipóósk I danced
nitápoosk I am dancing
19) okát sleep
nitsó’k I slept
nitáioka’ I am sleeping
20) okstaki read+
áakstaki he’s reading
nitákstaki I am reading
21) waawahkaa walk+
áwaka he’s walking
Also, ‘aak-’ appears to mark the future, as in will or is going to. If we compare the verb forms in (22), we can see that it also affects pitch. In the present náóhkiwa he barks, the pitch is on the first syllable but when the ‘aak-’ marker is added the pitch shifts to it: ákaohkiwa he’s going to bark. This is also evident in examples (23) and (24).
22) ohki bark+
náóhkiwa he barks
áakaohkiwa he’s going to bark
áóhkiwa he’s barking
23) ohpotaa snow+
iihpotááwa it snowed+
áakohpotaayiihka it will snow+
áohpotaawa it is snowing+
24) otakó evening+
áakotakowa it will be evening+
25) iksistsikó day
áaksiksistsikowa it will be day+
I will mention that Frantz has both the durative and future markers listed with inherent pitch. As I never got a chance to hear these forms in isolation, I have chosen to list them as unspecified for pitch.
Additionally, we see that pitch shifts to the person marker in some paradigms, such as example (25). If we look at the third person, where no prefix is added, we see that pitch in on the third syllable. When the first or second person markers are affixed, the pitch shifts to them.
25) áksskini to know
nítsskini I know
kítsskini you know
isskinímiowa they know
Prefixes
Possession: ‘nit’ is the first person prefix, ‘kit’ is the second person prefix and ‘ot’ is the third person prefix
26) apayín bread
nitápayin my bread
kitápayin your bread
otápayin his/her bread
27) ottsís sausage
nitóttsim my sausage
kitóttsim your sausage
otóttsim his/her sausage
28) akssín bed
nitókssini my bed
kitókssini your bed
otóssini his/her bed
29) miistsís stick
nitsíístim my stick
kitsíístim your stick
otsíístim his/her stick
30) aohkíí water
nitáóhkiim my water
31) simssín drink
nisímssin my drink
kisímssin your drink
32) siksikimí tea
kitáísiksikimi your tea
áísiksikimi his/her tea
As these compounds illustrate, pitch that occurs on the second or third syllable of the noun shifts left to the morpheme boundary when combined with the possession prefixes. This is akin to the compound noun pitch shifts as these too shift to the leftmost syllable, but in this case the pitch stays on the rightmost morpheme of the morpheme boundary. For example, in (30) the pitch in aohkíí water shifts to the first syllable in the noun to become nitáóhkiim my water.
This also looks to be driven by weight sensitivity. This is less obvious in these possessive forms than in the compound nouns because the person marker is realized as ‘ni-’ or as ‘nit-’ (depending on whether the noun is vowel initial or not), changing syllabification and, hence, weight. For example, (26) apayín bread when prefixed by the first person marker becomes nitápayin my bread and would syllabify as /ni.táp.ay.in/. If it were to be syllabified as /nit.ap.ay.in/ we would expect pitch on the heavy syllable - the first person morpheme. Likewise in (31), simssín drink, is prefixed by ‘ni-’ and results in nisímssin my drink. If it were prefixed with nit- we would again expect pitch on the first person morpheme, as we see in the verb paradigm in (25).
Exceptions
I have also run into what appears to be exceptions to the general patterns I have illustrated so far, for example, (33).
33) isttókimaa drum
nitsisttókimaa my drum
kitsisttókimaa your drum
otsisttókimaa his/her drum
In this possessive paradigm, I would expect pitch to shift to the morpheme boundary when combined with the possession markers and give results such as: nitsísttokimaa my drum, kitsísttokimaa your drum and otsísttokimaa his/her drum, as the accented syllable would be heavy. Why pitch is not shifting to the morpheme boundary is puzzling and requires further analysis. There are more examples of this, as in (34).
34) ksikkokóówa tent
niksikkokóówa my tent
kiksikkokóówa your tent
oksikkokóówa his tent
Orthography
Blackfoot, for the most part, is a spoken language and there was no written system for it until the 1970s when D. G. Frantz created one (Kaneko: 1994). In much Blackfoot writing, pitch accent is not orthographically recorded, but Frantz did record this in his dictionary and grammar. I have adopted his transcription method by indicating pitch with an acute accent ‘ ´ ’ diacritic over the appropriate vowels.
I believe this is useful because pitch functions in two ways in the language. First, it discriminates between different lexical meanings. This use of pitch is completely unpredictable and, therefore, needs to be indicated in the orthography. Secondly, pitch is part of the phonology. For the most part, this use of pitch is predictable, therefore there is no explicit need to indicate it in the orthography. But, as I have briefly mentioned, there are exceptions to the patterns, and these exceptions do need to be noted explicitly. Other situations also occur where a compound word can have multiple pitch peaks (see Kaneko: 1994), and the use of the pitch diacritic clearly illustrates when a compound adopts the pitch of one of the morphemes and when it retains both.
Intonation
Another way of evaluating pitch is to examine how it interacts with the intonational patterns of the language. Intonation is realized as a tonal melody that occurs across an utterance (as opposed to pitch accent, which occurs at the word level) and is commonly used cross-linguistically to express emotion as well as playing many syntactic and pragmatic roles.
In general, as we say an utterance our pitch range continually drops so that a high pitch at the end of an utterance may be realized as phonetically lower than a low pitch at the beginning of an utterance; this is called declination. At this point I am unable to evaluate intonation, but wanted to note it is claimed that some words in Blackfoot have pitch accent even when no pitch peak is visible on a spectrogram (Van Der Mark, personal communication).
Examples
for this can be seen in spectrograms for two words I elicited:
pinááp east, and
kokóto ice (both of which are listed in Frantz (1989) as having
pitch). As the spectrograms show, there is no significant rise in pitch on the
accented vowel, rather the word as a whole shows a level pitch. Since we expect
the pitch to drop on the final vowel in kokóto ice, for example, and it
doesn’t, it indicates that something is preventing normal declination from
occurring. This is attributed to the relative high pitch of pitch accent.
Pitch for pinááp east /ináápo/
/i n á p i/
We can compare the acoustic pitch for
pinááp east with that of míín berry, which shows a visible rise in
pitch followed by a drop.
Pitch for míín berry /mííni/
/m íí n i/
Notes
Data followed by an + are taken from Frantz
Data followed by a * are taken from Kaneko
All other data was elicited for this project
References
Frantz, D. & Russell, N. 1989. Blackfoot dictionary of stems, roots, and affixes. Toronto: University of Toronto Press.
Kaneko, I. 1994. A metrical analysis of Blackfoot nominal accent in optimality theory. MA thesis: University of British Colombia.
Van Der Mark, S. 2001. Acoustic correlates of Blackfoot prominence. Honours Thesis: University of Calgary.