Stress Assignment in Yimas

Stress in Yimas is predictable. Namely, it assigns primary stress to the first syllable of the word, and secondary stress to the third syllable if the word consists of more than three syllables.  The rules are stated below:

However, an exception occurs when the first vowel of the word is epenthetic. In this case, the epenthetic vowel is not stressed, and primary stress moves to the second vowel in the word.  Furthermore, these forms never bear secondary stress.

Moreover, there is an additional exception when the first two vowels of a word are epenthetic: the Primary Stress Rule (1) takes place (the first syllable receives primary stress).  This is because one of the first two vowels must bear primary stress.

When affixation occurs, we can see that the addition of a morpheme affects the assignment of stress.  An affix always receives a particular stress assignment—it can either be stressed or unstressed.  The following example outlines what occurs when a morpheme that always receives secondary stress is added: 

In order to create the dual (DL) and plural (PL) forms of the word kɨpáŋ “leg”, the morphemes /-kɨl/ and /-kɨ/ are suffixed, respectively.  These morphemes always receive secondary stress; thus, in kɨpáŋkɨ̀l and kɨpáŋkɨ̀, primary and secondary stress are in adjacent syllables.  According to the Epenthesis Stress Rule (3) outlined above, this should not happen—only primary stress should exist. Moreover, with the consideration of the Primary and Secondary Stress Rules (1) and (2), there is evidence that in Yimas it is more desirable that stress alternate syllables.  Because the final form of the word violates rules (1), (2), and (3), one can conclude that stress assignment does not occur all at once; it in fact in cycles.

Example (5) shows stress application if /kpaŋkɨl/ were monomorphemic. The Epenthesis Stress Rule (3) would generate primary stress on the second syllable because the first vowel is epenthesized.  Consequently, the incorrect form *kɨpáŋkɨl is produced.

In Example (6), we see that stress is not applied to the entire word as a whole because this form truly consists of two morphemes, corresponding to two different cycles.  First, the Epenthesis Stress Rule (3) applies to the root /kpaŋ/ in Cycle 1, resulting in kɨpáŋ. Then in Cycle 2, the suffix /-kɨl/ is added to the root. This morpheme always receives secondary stress, regardless of position, resulting in the correct form: kɨpáŋkɨ̀l.

Lastly, an example of the addition of a type of morpheme that is always unstressed is presented below:

Pronominal prefixes in Yimas are normally unstressed, so the stress rules (1), (2), (3), and (4) are applied to the root as if there were no prefixal syllables.  This is also evidence of cycles.

Example (8) shows stress application if there were only one cycle.  Primary and Secondary Stress Rules (1) and (2) assign stress on the first and third vowel.  This produces the incorrect form *pía-kà-tɨmi.

Example (9) shows stress application in two cycles.  First, stress is applied to the verbal root.  Because a vowel is epenthesized, Epenthesis Stress Rule (3) generates tɨmí initially.  Next, the two pronominal prefixes are added in Cycle 2.  They never receive stress, so no stress is added in this cycle, and the form produced in Cycle 1 is preserved.

---

Foley, W. (1991). Phonology. In The Yimas Language of New Guinea (pp. 75-80). Stanford, California: Stanford University Press.

 

Noun Incorporation in Yimas

According to Mark C. Baker, Noun Incorporation structures are “those in which a nominal root is morphologically combined with a predicate to make a single, complex word.” This process is common in many polysynthetic languages. In Yimas, however, Noun Incorporation is relatively rare – it usually generates idiomatic structures, except in the cases where nominalized adjectives and adjectival verbs are incorporated.

An example of Noun Incorporation in which the resulting structure is idiom-like in meaning is illustrated below:

(1) ampan-pay-pucm-api-n

     HORT 1A/2SG D – first – time VII SG – put in – IMP

     “I will give you time first.” (I will wait)

 

The bolded segments in example (1) indicate where Noun Incorporation is occurring. The resulting utterance is not completely transparent in meaning, as is typical of this process in Yimas. As a result, Noun Incorporation is not considered as productive in Yimas as it is in other polysynthetic languages.

However, there are instances of Noun Incorporation that yield more compositional, hence more literal, utterances:

(2) patn                        na-mpu-ŋa-taŋ-

     betelnut V SG     V SG O – 3PL A – 1SG D – COM –

     mama–k–n-tal-cɨ-t

     bad – IRR – V SG – CAUS – become – PERF

     “They ruined my betelnut.”

In example (2), the nominalized adjective mama-k-n “bad” is incorporated into the causative verb. The result is more transparent than example (1) in that the connection between “to make bad” and “to ruin” is more apparent than the connection between “to give time” and “to wait.” Although this resembles causativization of an adjective, like the example [_Amodern[_Vize]] shown in class, it is explicitly cited as an example of Noun Incorporation in Yimas by Foley.

It is also important to note that in Yimas Noun Incorporation is not required; therefore, nouns are not restricted to incorporated positions:

(3) ura-mpu-na-akpi-api-n

      fire O – 3PL A – DEF – back V SG – put in – PRES

     “They are putting (their) backs to the fire.”

(4) akpi                ura-mpu-na-api-n

       back V SG      fire O – 3PL A – DEF – put inside – PRES

      “They are putting (their) backs to the fire.”

Examples (3) and (4) show two different constructions of the same sentence. In example (3) the noun akpi “back” is incorporated into the verb structure, whereas in (4) it stands alone. Because Noun Incorporation is optional in Yimas, these two sentences have identical meanings.

The tree below, corresponding to example (2), further illustrates the process of Noun Incorporation in Yimas: 

Within the tree, we see the BPC used for two separate processes. First, the node containing the adverb pay (“first”), undergoes adjunction-modification, and thus does not percolate any additional features onto the verb. Secondly, the BPC is used to percolate inflectional morphemes in two instances: the hortative, (“HORT”) and imperative, -n (“Imperative”). Moreover, our tree demonstrates the use of a portmanteau to form the complex noun morpheme, ampan. Syntactically, this node would include agreement structures for the subject (HORT 1A) and indirect object (2SG D) that combine together to form the resulting noun morpheme. Finally, the tree illustrates the fact that the agreement structures within ampan do not saturate; and therefore, there is no linking occurring between ampan and the existing verb phrase “put in time”.

References:

Baker, M. (1993). Noun incorporation and the nature of linguistic representation. In Foley, W. A., editor, The Role of Theory in Language Description, pages 13–44. Mouton de Gruyter., Berlin.

Foley, W. (1991). Basic Verbal Morphology. In The Yimas Language of New Guinea (pp. 319-321). Stanford, California: Stanford University Press.

Inflection of Yimas

The Yimas language is spoken by the Yimas peoples of New Guinea. According to data collected in 2000, there are approximately 300 native speakers. Hence, this language is considered potentially endangered.

Yimas has inflectional prefixes that encode subject and object agreement. The following sentences are simple examples (Note: Foley uses A to represent the subject of a transitive verb and S to represent the subject of an intransitive verb. This distinction exists because Yimas is an ergative language.):

(1). pu-n-tay

3PL O—3SG A—see

‘He saw them.’

(2). na-mpu-tay

3SG O—3PL A—see

‘They saw him.’

In addition to demonstrating the inflectional prefixes, these examples also exhibit an interesting phenomenon with respect to the order of subject and object agreement morphemes. Specifically, object agreement occurs further away from the root than does subject agreement, which is very rare cross-linguistically.

Below is a classification of the inflectional prefixes used in (1) and (2).

Morpheme	Agreement Marker	Meaning
na-	3rd person singular object	Him/her
n-	3rd person singular subject	He/she
pu-	3rd person plural object	Them
mpu-	3rd person plural subject	They

The relationships between these morphemes are clear. The third person singular object and subject have /n/ in common. The third person plural object and subject have /pu/ in common.

Furthermore, Yimas is known for having a considerably free word order due to its inflectional prefixes. An example of this observation appears in the pair of sentences below:

(3).  payum     narmaŋ            na-mpu-tay

men           woman             3SG O—3PL S—see

‘The men saw the woman.’

(4).   payum       narmaŋ             pu-n-tay

men           woman             3PL O—3SG S—see

‘The woman saw the men.’

As you can see, the lexical entries for ‘men’ and ‘woman’ occur in the same order in these sentences. This is acceptable in Yimas because the inflectional prefixes that attach to the verb indicate which lexical entry is the subject and which is the object. However, when the subject and object do not differ in number the meaning of the verb becomes ambiguous. In this case, the word order would be more restricted.

As each inflectional affix is added to the root, it then adds its own features to the V node that dominates it. Thus, the Word Structure Tree for example (1) is as follows:

Foley, W. (1991). Basic Verbal Morphology. In The Yimas Language of New Guinea (pp. 193-195). Stanford, California: Stanford University Press.