A Phonetic Analysis of Blackfoot

by Jeff Long

[Corrections by Darin Howe]

            The language of Blackfoot has many phonemes [relatively few, actually] and here I plan to investigate and to show with supporting data exactly what phonemes this Algonquian language has. Although more data is essentially required, this analysis is adequate in a first attempt at capturing the sounds used in Blackfoot [although I asked that you discover Blackfoot phonetics on your own, you need to mention previous descriptions, especially by Taylor and Frantz]. I will discuss the phonemes based on the IPA and explain for example, their manner of articulation, place of articulation and I will also provide the accompanying sound files based on my word list elicited from our language expert Mrs. Rachel Ermineskin.

Consonants -

            First off, we look at the plosives, or stops of the language and find that [p],[t] and [k] exist with an occasional voiced counterpart occurring in instances between vowels.

Starting with the phoneme [p] [use /p/], this plosive is bilabial and appears to unpredictable and distinctive in that it occurs word initially, medially, and finally:

[pʊkun] [pʊkún], [kipu] [ki:pó], and [ kajskap] [ kajʔskáxp]. [provide Blackfoot orthography & glosses: pokón 'ball', kiipó 'ten', kai'skááhp 'porcupine'.]

Although it doesn’t seem that the phoneme [b] /b/ occurs, an allophone [b] of the phoneme [p] /p/ tends to occur between vowels; this is the voiced bilabial plosive and in this case is predictable and will often occur between vowels. This is most likely due to voicing assimilation provided by the environment of the vowels and causes occasional voicing of the normally voiceless [p] phoneme: [ʌmskabibigani] [aamsskáápipikani 'Peigan tribe'; you'll need to provide a spectrogram to back up your claim here, as these stops appear to be simply unaspirated.]

Note that the phoneme [p] is unaspirated and this becomes different quite quickly for English speakers, as we tend to aspirate the phoneme [p] quite often. This can make it difficult to hear an unaspirated [p] as voiceless and tends to be heard by English speakers as [b] even though it remains voiceless and unaspirated [p]. These differences however are shown with spectrogram analysis very clearly with acoustic phonetics. In this case I suggest looking at the acoustic phonetics section for further analysis.

            The next phoneme [t] we look at is a voiceless alveolar plosive, and like the [p] phoneme occurs in an unpredictable distribution thus making it a phoneme and not an allophone. Examples include: [tʌka:], [imita], and [mistakist]. Interesting to this phoneme is that aspiration will occur sometimes before a vowel, but it seems to be in free variation in that it does not always occur: [nit^hukska], and [nisitu]. Another point worth mentioning is that the voiced alveolar flap [ɾ] can occur between vowels similar to [b] occurring instead of [p] between vowels as mentioned above. The key difference is that the flap occurs much quicker as the tongue tip quickly touches the roof of the mouth much like alveolar [d], only faster. This is similar to the way we say ‘water’ in English, not pronouncing a full [d] sound, but a quicker version, namely the voiced alveolar flap [ɾ], a sound between [t] and [d].

            Next, we examine the [k] sound, or plosive voiceless velar phoneme. Based on the data this quickly turns out to be unpredictable and therefore a phoneme with the occasional voiced allophone [g] occurring between vowels: [k] in the name for ‘South Peigan’ [ʌmskabibikani] and [g] in the name for ‘Peigan Tribe’ [ʌmskabibigani]. Here it is again worth noting that the vowels before and after most likely play an influential role on the voicing of [k] to [g] and this does not occur anywhere else or very often. For that matter this makes [g] an allophone of [k]. To show some examples of [k], we look at the words ‘ball’, and ‘hot’: [pokun], and [kipu]. In my data I have no examples of [k] appearing at the end of a word which is curious and makes it worth searching for in future elicitations.

            Next, we look at the nasals to see what place features they have in Blackfoot. The bilabial [m] does in fact exist and seems to only have the form [m] making this the only allophone of the phoneme. There is however an instance of lengthening of the [m] in the word laugh as shown: [aim:i]. Because this is the only instance of the lengthened [m] in the data its not for sure an allophone, and more instances of a lengthened [m] are to be looked for. [m] does occur between two vowels such as in ‘coffee’: [ɔxkuɾusajskskimi:] and in this instance it is not lengthened. Also, in the word ‘dog’: [imita] which has two [i] vowels such as ‘to laugh’ or [aim:i], ruling out the idea that two high vowels cause vowel length. [m] also occurs word initially such as in ‘foot’ - [mɔxkajt^si] and word finally such as ‘medicine’- [sa:m].

            Next we look at the alveolar /n/ nasal phoneme and notice that it has the allophones alveolar [n] and palatal [ ɲ] .[n] occurs word initially, word medially and word finally: [nanisu:], [akani], and [oxkinan]. This is unpredictable and therefore phonemic whereas [ ɲ] only occurs before the glide [ j ] which is predictable and therefore an allophone of the phoneme [n]. Examples include: ‘#3’ [ ɲjoskani:], and  ‘really’- [ʌɲja] and in both cases we see the [ ɲ] followed by the glide. Nowhere else in the data shows this therefore based on what we have, it appears the  [ ɲ] remains a predictable allophone of [n].

            Before going further, it is important to step back a moment and look at the glottal stop [ʔ]: This is a plosive sound and although it doesn’t occur frequently in my data, it is there, and we must look at it’s environment. There is a very clear example of the glottal stop [ʔ] in the word ‘young’- [aʔsitapi] following a vowel and preceding a fricative. Also occurring is the glottal stop between the two [a] vowels in the word for ‘Peigan and Siksika First Nations’- [pikania^ʔan]. This seems to break up the [a] sound and allow for another syllable. Also interesting is that the [a] has a glottal stop before moving into the [s] in the above word for ‘young’ making a clear break in transition from [a] to [s]. The data isn’t extensive here, but it demonstrates the glottal stop very clearly and thus makes it valid in our phonetic inventory.

            Turning now to trills, we quickly notice that they are completely absent from all the data. This is fascinating because it is a manner of articulation not needed in Blackfoot, and therefore not used. This is not to say that a Blackfoot speaker cannot make these sounds i.e. such as a bilabial [β] or an alveolar trill [r], as they may use them for effect when telling stories or using onomatopoeia. A comparison in English is that although we don’t normally use a trilled [r], this sound can be made quite easily by many speakers and may be used for any particular reason, the same being true in Blackfoot.

            Moving now to a very intriguing aspect of the Blackfoot phonetics, we examine the fricatives. Firstly, we see the alveolar fricative [s] occurring in an unpredictable distribution therefore making it a phoneme. Examples include: [moxksɪsɪs], [sa:m] and [kokit^siksis].  Note, there isn’t any voiced version of [s], which would be [z] anywhere in the data.

The [ks] cluster appears to be an affricate (co-articulation) of phonemes [k] and [s] but is most likely two separate consonants. For example, [ɛkst^sɪksi], or ‘to scratch’ appears to have a [ks] affricate but is most likely only two separate consonants in both instances. Also in the word for ‘nose’ [moxksɪsɪs], based on Krista Jackson’s syllable structure in the ‘prosodic phonology’ section of this website, we see that only two consonants could be allowed to be in the coda position and it makes the most phonetic sense to have a velar [k] follow a velar [x]. Therefore we assume that [xk] is the coda and [s] begins the onset of the next syllable. Leaving that section alone, and staying focused on phonetics we conclude that [ks] is not an affricate.

            Next, we see something very interesting happening with the voiceless palatal, and velar fricatives, [ç] and [x], respectively. It seems that the phoneme [x] occurs after low and mid vowels, [oxkinan], [kajskaxp]; word initially and before velar [k], [xkɪt^sika], [xini:]; and word medially, [inɛsp^hxkṵ], [mistaksupxkxk^hina] and therefore proves to be a phoneme. The allophone [ç] on the other hand only appears after the high vowel [i] and therefore is very predictable and it becomes obvious that this is an allophone of [x], in complementary distribution, that is never occurring where a velar [x] would. Examples are shown in the words: ‘can opener’ - [içtɛkainkiup], ‘radio’ - [iÇtaxt^simop], and ‘worried’ - [hʌpejt^siÇta]. In all three instances the palatal fricative [ç] follows the high front [i] vowel. Something very intriguing is that I have no data with a high back [u] vowel preceding a fricative. This might be expected to be [x] due to lip rounding and the backness of the [u] vowel, but it is worth noting and watching for in future research.

            There is also the fricative [h] occurring in words such as ‘worried’ - [hʌpejt^siÇta], but it seems to be rare or at least in this data set. Nevertheless it occurs and therefore is a phoneme in the Blackfoot language. More research will be required but this example makes it very clear the [h] is present.

            Finally we look at the glide [ j] , or more specifically, the palatal approximant [ j]. This is a phoneme that interacts quite frequently with vowels in producing diphthongs, but there are examples that show it to be present as it’s own phoneme as well. An excellent example of this is the word for ‘wolf howl’ - [ajat^hu]. Here, it is clear that the sound sequence is [a-j-a] and it is not the diphthong [aj] such as in the Canadian English word ‘white’ or ‘bike’. Equally notable is that in the word ‘wolf howl’, the length of time for the vowels being released is equal to the length of time the glide is released showing it to be it’s own phoneme. Acoustic phonetics is better in explaining this with a spectrogram so for now we move on.

            There are no alveolar approximants or lateral approximants in the data and therefore they are not included in our phonetic analysis of Blackfoot. Equally noted is that lateral fricatives do not exist either, based on the data elicited.

A quick summary of the consonants of Blackfoot thus far based on the IPA is:

Vowels-

            Based on the data in the elicitation, there appears to be 11 vowels in Blackfoot if we include the schwa, but based on the data it appears that the schwa doesn’t truly exist. Note that for all vowels, the various positions of the lips and tongue give the vowels their different and unique qualities. The following chart gives us an idea of how these vowels are related to each other and yet distinguishable:

Front           Central            Back

High             i                                                  u

                              I                                 ʊ

Mid                    e                     ə                      o

                              ɛ                                ʌ  ɔ

Low                         a

Before moving on, we look at an interesting feature about the [u] and [o] vowel. Our language advisor Rachel mentioned that in Siksika Blackfoot the word ‘ball’- [pokun] has more of an [o] sound whereas another dialect, namely the Peigan dialect uses more of the [u] sound:  ‘ball’- [p^hukun]. She indicates that the Peigan dialect uses more aspiration as well as I’ve shown with the transcription. To further demonstrate this comparison, she compares the initial [po] and [pu] syllables and their sounds, [po] being the Siksika dialect and [pu], the Peigan dialect: [po] vrs [pu]. It is worth noting that these sounds are in fact quite similar but yet Rachel can notice a distinctive difference very easily between the two dialects. This tells us something important about the dialects of Blackfoot in that they have some slight yet distinguishable vowel differences. Based on the dictionary writing which spells the word ‘ball’ as pokon, and then hearing Rachel say the word, the question of a free variation between [o] and [u] arises. Looking more extensively at the data and speaking with Rachel our co-investigator, it becomes apparent that these are indeed two separate phonemes. Examples of words with [o] help to clarify this: ‘sunrise’ – [otamiskapinatusi], ‘ankle’ – [oxkinan], ‘wise – [mokakʲi]. A question that remains is if one used the [u] vowel instead of [o] in the environments above, would it sound like another dialect, or altogether ungrammatical. More research will likely uncover this.

Dipthongs-

            Blackfoot also possesses diphthongs, which are in essence the glide [ j] combined with a vowel, such as [aj] e.g. ‘dirty or messy person’ - [skaisajɪnskakapsi] or [ jo] e.g.  ‘#3’ – [ ɲjoskani:] . A rising diphthong occurs when the greatest intensity is found on the final part of the vowel combination. An example of this would be ‘really!’- [ʌɲja], and again, ‘#3’ - [ ɲjoskani:]. A falling diphthong as then would be expected is when the primary part of the diphthong contains the greatest intensity, and the intensity trails away or ‘falls off’ as it proceeds. Examples include: ‘#6’ - [nɔ:j], ‘worried’ - [hʌpejt^siÇta], ‘foot’ - [mɔxkajt^si], and ‘North Peigan’ - [apʌt^wxsipikajni̥]. It’s again important to note that the length of the glide [ j ] here is much shorter than the vowel. As noted above, a pure glide such as in ‘wolf howl’ - [ajat^hu], the glide has equal the release length of the [a] vowel. In true diphthongs this length is greatly reduced. In the data it is quite difficult at first to tell if the word for ‘Blood Tribe’ is a diphthong [akajna] or has the two vowels [a] and [i] as in [akaina]. After some acoustic analysis using the program Praat, and looking at vowel length very closely, it becomes apparent that these are in fact two distinct vowels, rather than the [a] vowel and the reduced glide [ j]. 

Vowel Lengthening-

            As shown by the transcriptions, some of the vowels tend to lengthen. This means that they are longer than the average vowel, and this was investigated further by using acoustic analysis. In essence, the duration was measured to be approximately twice as long as a regular vowel and this lengthening was observed in different vowels. Some examples based on the data are shown below:

            It appears that there is a lot of vowel lengthening at the end of words, yet there are also regular vowels found at the ends of words. Therefore, the phonological patterning of Blackfoot would better be able to account for this., although is it important to make note of this in describing the phonetics of Blackfoot.

Devoiced Vowels-

Equally notable as vowel lengthening, is that there is devoiced vowels. Examples are shown with:

‘brain’ – [up^hi̥], ‘Wednesday’ – umxʧoskasikatuikst^siku̥pi̥], and ‘North Peigan (Canada)’ – [apʌt^wxsipikajni̥] to name a few. Here we see not only the front high [i] vowel being devoiced, but the back high vowel [u] as well. This again will be better captured in terms of phonology, but is important again to note upon, as these are clear examples of devoiced vowels. Another possibility could be that perhaps our research consultant was ‘trailing off’ so to speak at the end of the words. Even still, we have to find out if they are in fact always unvoiced, or due to other circumstances.

Affricates-

            Based on the data, it seems that there exists one affricate, or double articulation and this is the [t^s] affricate. This appears to occur only in front of high vowels, however there are some interesting examples to mention:

‘grasshopper’ – [t^sɪkat^si], ‘chocolate’ – [opaki:t^ski], ‘foot’ – [mɔxkajt^si],

‘radio’ – [iÇtaxt^simop], ‘scruffy or unshaven’ – [ist^st^sui], ‘#7’ – [xkɪt^sika],

‘stick’ – [mist^sis], ‘to scratch’ – [ɛkst^sɪksi], ‘toes’ - [kokit^sis], ‘where’ – [t^sima:],

‘you don’t say’ - [kɪt^simaniÇp]

As we can see this affricate is quite prevalent throughout the language. It does appear in front of the velar plosive [k] however in ‘chocolate’ which is curious and not only in front of high vowels as we see in ‘grasshopper’ and ‘to scratch’. Also interesting is the apparent gemination of this [t^s] affricate in the word ‘scruffy or unshaven’.  The fact that we see it in front of [u] supports the idea of it preceding high vowels, but more research and data will needed, and again, perhaps phonology can help to shed light upon as to what causes this [t^s] double articulation to occur. Nonetheless, we see numerous examples here which is beneficial to learning and understanding the phonetics of Blackfoot.

            On another note, Our language expert Rachel noted that when the English names Charlie and Johnny are used, they both are changed to [t^sani]. The voiced version of [ʤ] in the English version [ʤani] changes to the affricate [t^s], in [t^sani] and in some instances the affricate [ʧ], as in [ʧani]. This [ʧ] affricate is the ‘ch’ sound at the beginning of the English words ‘Charlie’ or ‘chance’.

Examples are shown with some of the data elicited by Chirlene Lee:

‘John wants that hat’; in this example we hear the [t^s] affricate in [t^sani]

‘John wants those (many) hats’; in this example we hear the [ʧ] as in [ʧani].

In the second example, it is difficult to distinguish whether or not voicing is there as in [ʤani], but acoustic analysis reveals voicelessness. This makes sense because as claimed above the only close allophone of [t] is the flap [ɾ], which isn’t used in double articulations.

            What is perhaps the most interesting about these name changes into Blackfoot is the fact that in the word for ‘Charlie’, there is no [ɹ] approximant. Although Blackfoot

doesn’t use these sounds, it becomes interesting that they substitute the alveolar nasal [n] in it’s place. This is most likely because this way, the place feature alveolar is preserved and well as the fact that both sounds can be prolonged. For example, the alveolar approximant [ɹ] can be prolonged and is in essence a sonorant; the same being true with the alveolar nasal [n]. On the other hand the voiced alveolar flap [ɾ] is not prolonged at all. The only other option for place of articulation is the fricative [s], a ‘noisy’ obstruent which is simply not considered. This is a fascinating phonetic process occurring in Blackfoot with borrowed words. Rachel is able to use the name John exactly as it occurs in English, however she is a fluent speaker of English and therefore simply switching to English when it is pronounced this way.

Conclusion-

            Looking back at the phonetics of Blackfoot, we see a multitude of fascinating features. The interaction of the consonants and vowels, for example: the plosive consonants with the voiced allophones [b] and [g] being occasionally used for [p] and [k], but the lack of [d] being used for [t]; palatal fricative [ç] following the high front [i] vowel and the velar fricative [x] after low and mid vowels; the distinguishable use of glides and diphthongs; vowel lengthening and devoicing; affrication which appears to be influenced by high vowels following a [t], but not for certain; and finally the transformation of borrowed words from English into Blackfoot. These features give a good general overview of some of the key phonetic elements of the language worth noting and investigating. Although more research will need to be conducted, this provides an adequate headstart and covers some fundamental issues of phonetics as a whole. I look forward to learning more about this language based on the work of others on this website as well as on what future field work can help us to discover.

Blackfoot Word List:

Ankle – [oxkinan]   

Ball (Siksika Blackfoot) – [pʊkun] or [pukun] or even [pokun]

Ball (Peigan) – [p^hukun]

Blood Tribe – [akaina]

Brain – [up^hi̥]

Buffalo Jump – [inɛsp^hxkṵ]

Can Opener – [içtɛkainkiup]

Chocolate – [opaki:t^ski]

Coffee – [ɔxkuɾusajskskimi:]

Dirty / Messy Person – [skaisajɪnskakapsi]

Dog – [imita]

Egg – [uwe]

Energetic / Helpful – [iskunatapsi̥]

Enjoy – [eta:msi]

Foot – [mɔxkajt^si]

Grasshopper – [t^sɪkat^si]

Hot – [ksistui:]

Medicine – [sa:m]

Mountain – [mistakist]

Mountain Goat – [mistaksupxkxk^hina]

North Peigan (Canada) – [apʌt^wxsipikajni̥]

Nose – [moxksɪsɪs]

#8 – [nanisu:]

#5 – [nisitu]

#4 – [nisuɔ:]

 #1 – [nit^hukska]

#6 – [nɔ:j]

#10 – [kipu]

#3 – [ ɲjoskani:]

Peigan + Siksika First Nations – [pikania^ʔan]

Peigan Tribe – [ʌmskabibigani]

Porcupine – [kajskaxp] or [ kajskap]

Radio – [iÇtaxt^simop]

Really! (Are you telling the truth?) – [ʌɲja]

Scruffy / Unshaven – [ist^st^sui]

#7 – [xkɪt^sika]

Siksika – [sɪksika]

Sneaky – [ɛsimjapsi]

South Peigan – [ʌmskabibikani]

Stick – [mist^sis]

Sunrise – [otamiskapinatusi]

‘They are saying’ – [awanija]

To Laugh – [aim:i]

To Scratch – [ɛkst^sɪksi]

Toes – [kokit^siksis] or [kokit^sis]

Tomato – [xini:]

#2 – [nat^huka]

Wednesday – umxʧoskasikatuikst^siku̥pi̥]

Where – [t^sima:]

Who – [tʌka:]

Wise – [mokakʲi]

Wolf Howl – [ajat^hu]

Worried – [hʌpejt^siÇta]

You don’t say (Are you telling the truth?) – [kət^simaniÇp] or [kɪt^simaniÇp]

You Hurt (you are injured) – [akaxsi]

You Hurt (you injure someone else) – [akani]

Young – [aʔsitapi]

References-

Most importantly I would like to acknowledge our Blackfoot language expert Mrs. Rachel Ermineskin, with whom made all of this possible.

Ball, Martin J. & Joan Rahilly. 1999. Phonetics The Science Of Speech, New York;

Oxford University Press Inc.

Dalbor, John B. 1997. Spanish Pronunciation Theory And Practice (Third Edition),

Philadelphia; Harcourt College Publishers.

Fromkin, V., Hultin, N., Logan, H., & R. Rodman. 1997. An Introduction To Language

(First Canadian Edition), Toronto; Harcourt Canada Ltd.