The Blackfoot Syllable analyzed in the Optimality Theory

By Krista Jackson

Syllables consist of an onset, nucleus, and coda.  The onset and coda are occupied by consonants forming the margins of the syllable, and the nucleus is universally obligatory and occupied by vowels.  Thus, the universal structure for syllables is:

Onset

            The onset in the Blackfoot syllable is optional, however, it can only occupy one skeletal slot with the exception of /s/.  When /s/ occupies the first skeletal slot, another consonant may follow it.  However, this will be treated as a special case.  Similarly, English and Dutch treat /s/ as a special case when more than two consonants occupy the onset position. Therefore, onsets are optional; Onset (“Syllables must have onsets.”) (Kager, 93) ranks low and complex onsets are not permitted unless the first consonant is an /s/; *ComplexONS (“Onsets must be simple”) (Kager, 97) ranks high.

In order to correspond to the Blackfoot syllable template (that does not allow complex onsets), epenthesis or deletion may be used.  Blackfoot commonly uses both epenthesis and deletion to adhere to the markedness constraint that disallows complex onsets.  Thus, DEP-IO (“Output segments must have input correspondents”) (Kager, 101) ranks low and MAX-IO (“Input segments must have output correspondents”) (Kager, 102) rank high. 

Also, alignment must be considered here.  If a word is perfectly aligned, then the prosodic structure and the morphological structure should match perfectly.  Align-L (“The left edge of the Grammatical Word coincides with the left edge of the Prosodic Word.”) (Kager, 111) ranks low in Blackfoot.  Because epenthesized vowels may be at the beginning of the Prosodic word, Align-L is violated as the Grammatical Word does not perfectly align.  

*/ts/ is an affricate, one segment occupying one skeletal slot.

*Please note that the two violations for *ComplexONS that are not fatal to the word are the /s/ exceptions.

            Most all consonants found in the Blackfoot language (please refer to the section on Blackfoot sounds) can occupy the onset position except for a few, which are /h/, /w/, and /y/.  When asked to produce a word that started with /h/, Rachel could not think of any word that started with it.  Words beginning with /w/ or /y/ were produced with an epenthesized vowel inserted immediately before it.

            There are two affricates that can also occupy the onset position, that are one segment taking the position of one skeletal slot.  These affricates are /ts/ and /ks/.  An example of the affricate /ts/ is above in the word /si.tsi.ka/ for ‘another’ and an example of the affricate /ks/ is in the word /a.wak.ksis/ for ‘chewing gum’.  It is important to note that the first /k/ belongs to the coda in the second syllable and /ks/ belongs to the onset in the last syllable.  Most languages consider /ts/ as common affricate but similar findings for /ks/ have not been found across languages.  However, /ks/ behaves identical to that of the affricate /ts/ in the Blackfoot language and should thus be considered an affricate in this language.

Rhyme

Nucleus

            The nucleus can occupy up to two skeletal slots.  There are three different vowels that can occur on there own (occupying one skeletal slot) in nucleus position, /a/, /i/, and /o/.  For example:

      A – IHTATSIKI – middle

I  – AKSSIN – bed

O – OTAHKOINATTSI - yellow

The nucleus can also occupy a maximum of two skeletal slots with either long vowels or diphthongs. For example:

            AA – KAAPOKSIINIMAAN – floor

            II – OHKIIMAAN – wife

            OO – AKOOPSKAAN – soup

      AO – SAO’TSI – remove/ take out

AI – IIHTAISINAAKIO’P – pencil

OI – OTAHKOINATTSI – yellow

            Words appearing to have three vowels in the nucleus are commonly separated by the placement of the stress.  For example, the word for ‘medicine’ /saa.am/ appears to have three vowels in the nucleus position.  However, when listened to closely it is apparent that the syllable boundary lies between the second and third /a/.

Coda

Like English, the coda is optional in Blackfoot.  Like the onset, the coda cannot be complex either.  Only one skeletal slot may be occupied in coda position.  Therefore, No-Coda (“Syllables are open.”) (Kager, 94) ranks low against faithfulness and *ComplexCOD (“Codas are simple.”) (Kager, 97) ranks high against faithfulness. 

            Geminates or long consonants are frequently seen in the Blackfoot language.  Normally, these geminates would occupy two skeletal slots in the onset or coda, thus producing a complex onset or coda.  However, epenthesis is used to avoid the complex margin if needed but otherwise, the syllable boundary lies in between the geminate forcing one consonant to belong to the coda of the first syllable and one consonant to belong to the onset of the second syllable.  Therefore, DEP-IO (“No epenthesis”) (Kager, 101) ranks low against faithfulness.  It also seems that at the end of words, deletion may also be used to avoid the complex coda.  Therefore, MAX-IO (“No deletion”) (Kager, 102) must also rank low against faithfulness.

            As shown above in the tables, it seems that either method of avoiding the complex coda may be used.  How the method is obtained is not yet clear, as further investigation into this matter is needed.  It can be said that epenthesis always “results in a decrease of structural markedness” (Kager, 105) or in other words epenthesis facilitates unmarked structures at the expense of faithfulness.  However, deletion also acts in this way as well.  Deletion takes place to eliminate the marked structure, complex coda, at the expense of faithfulness.

            “Economy, or against excessive epenthesis” (Kager, 104) proposes that epenthesis should be used to change the input “only under duress” or only when necessary.  Therefore, faithfulness to the input should only be violated when completely necessary; “faithfulness will always decide in favor of the most faithful candidate.” (Kager, 104)  In Blackfoot, economy seems to be followed as epenthesis is only used to reduce structurally marked segments such as complex margins.  However, Blackfoot allows codas so epenthetic segments added to avoid codas would be against economy restrictions.  Therefore, input faithfulness takes precedence over avoidance of codas and no complex margins takes precedence over no epenthesis:

            Just like alignment regarding the left side of the word, alignment must also be considered with the right side of the word.  If a word is perfectly aligned, then the prosodic structure and the morphological structure should match perfectly.  Align-R (“The right edge of the Grammatical Word coincides with the right edge of a syllable.”) (Kager, 113) ranks low in Blackfoot.  Because epenthesized vowels may be at the right edge of the Prosodic Word, Align-R is violated, as the Grammatical Word does not perfectly align.

Conclusion

            It seems to be clear that both the onset and coda cannot be complex.  Every example that I have obtained thus far is proof for my hypothesis.  However, further analysis of the Blackfoot syllable may be needed to support my claim.  The data that I have obtained thus far is only a sampling of the many words and syllables in the Blackfoot language.  Therefore, I propose that the Blackfoot syllable template is as follows:

Works Cited

Kager, Rene.  Optimality Theory.  Cambridge: 2001, Cambridge University Press