Kampa Second-position Clitics
[I intended to post this on June 27, 2017, but forgot to before I left internet contact.]
I am writing from Sepahua, one of the southernmost points in the Ucayali region of Peru. After a two-day delay, tomorrow I (hopefully) head upriver to the community of Kitepámpani to carry out eight weeks of linguistic fieldwork on Caquinte, an Arawak language of the Kampan branch of that family spoken by 300-400 individuals in the headwaters of the Mipaya and Pogeni rivers in the Cuzco and Junín regions of the country, respectively. I arrived in Pucallpa the afternoon of Monday, June 19, and was delayed there four nights waiting for a direct flight to Sepahua. (These are small 8-12-passenger planes that usually fly twice weekly. The alternative was to leave Wednesday on a flight bound for Atalaya, then spend some eight to nine hours on a boat on Thursday.) I had planned to head to Kitepámpani on Monday, June 26, again to be delayed due to the very legitimate reason that the family that regularly takes me from Nuevo Mundo, a Matsigenka community on the Urubamba river, to Kitepámpani, came to Sepahua for the birthday of their eldest daughter.
My dissertation has figured little into this blog, and I want to take this opportunity to change that somewhat. My dissertation focuses on the semantics and pragmatics of three second-position clitics in Caquinte -- =sa, =sakanika, and =te -- and their combinations (i.e., =sakanika=te). To a lesser extent it deals with two other second-position clitics -- =tari and =npa -- given that these frequently combine with =te. Despite the grammatical similarities across Kampan languages, the domain of second-position clitics constitutes a notable locus of variation. My dissertation is not comparative or historical in nature, but I am in many ways a historical linguist at heart, and I want to take a few paragraphs here to give some impression of this variation.
The semantics of second-position clitics in Kampan languages are subtle, often consisting of not-at-issue meanings, and as such they are one of the more difficult domains of study in these languages. In practical terms, they are often given very brief descriptions in unexpected sections of dictionaries and/or grammars. Elena Mihas' research on the Perené River dialect of Ashéninka -- in particular her 2015 grammar -- is the single most comprehensive effort to describe these elements. Here, however, I would like to focus on the two Kampan varieties with which Caquinte is presently and has been historically in most contact. I draw on my own field data from Caquinte, as well as early research on the Tambo River dialect of Asháninka conducted by Lee Kindberg (1980:461-466) and corpus-based research on Matsigenka conducted by Lev Michael and myself beginning in 2011 (based on research conducted by him and Christine Beier the same year in Quillabamba, Peru, in conjunction with speaker Haroldo Vargas Pereira). (See Vargas Pereira et al. 2013.) The aim is to give an approximate sense of variation among second-position clitics in these three varieties, based on the data presented in Figure 1. The table is organized based on presumed cognacy, with brief glosses or translations provided for all forms. This is not an exhaustive list of all second-position clitics in each language, nor are the analyses on which glosses and translations are based conclusive. There are almost certainly gaps in the data due to the nature of the sources on which this data is drawn. (In fact, data from Perené Ashéninka, for example, shows some forms cognate to those in Matsigenka.)
We see that only =me, =npa, =ta, and =tari are common to all three varieties. These clitics -- for other reasons not discussed here -- are probably reconstructable to proto-Kampan. Caquinte exhibits eight clitics not attested in the other varieties; Asháninka exhibits seven; and Matsigenka eight. In the context of my own dissertation, we see that Caquinte =sa and =te have cognates in Asháninka but not in Matsigenka, while =sakanika is unique to Caquinte -- the two former are described as markers of "impatience" in Asháninka. Other forms are the clear result of grammaticalization processes that lead to the detection of additional cognates. Caquinte, for example, exhibits three inflectable demonstratives ka, ra, and nta. The proximal demonstrative ka is likely the source of the relativizer =ka. Similarly, a cognate to the Caquinte demonstrative ra is likely the source of the Matsigenka relativizer =rira (-ri being a productive nominalizer in many Kampan varieties). (The Caquinte epistemic modal =ka has a different etymological source, and exhibits reflexes in Nanti (Michael 2008).) Elsewhere cognates are detectable only in morphologically frozen forms, e.g., Caquinte teekera(ta) 'not yet' (cf. dialectal variant teekerota).
Some "long" clitics in some languages are arguably etymologically related to segmentally shorter ones. Caquinte =sakanika is an example of this, although it is not obvious how it could be compositionally analyzed as consisting of =sa, =ka, and =nika. Other possible cases are =nika and =rika; Caquinte =satine and =shiatsi; Matsigenka =kario, =ratyo, =rira, =roro, =tata, and =ukari. I will not speculate about Asháninka etymologies.
I hope to someday carry out a team-based project focused on variation across Kampan languages and dialects. This sort of endeavor has at its core the study of lexical variation, but the morphologically incredibly rich nature of Kampan languages also demands careful comparative study of morphosyntactic differences in order to arrive at the proper understanding of the proto-Kampan verb phrase (second-position clitics included alongside innumerable verbal affixes), as well as issues related to inheritance, borrowing, and the like. Who is interested in fieldwork in the Manú, Urubamba, Apurimac, Ene, Tambo, Pichis, and/or Ucayali driver basins?
I am writing from Sepahua, one of the southernmost points in the Ucayali region of Peru. After a two-day delay, tomorrow I (hopefully) head upriver to the community of Kitepámpani to carry out eight weeks of linguistic fieldwork on Caquinte, an Arawak language of the Kampan branch of that family spoken by 300-400 individuals in the headwaters of the Mipaya and Pogeni rivers in the Cuzco and Junín regions of the country, respectively. I arrived in Pucallpa the afternoon of Monday, June 19, and was delayed there four nights waiting for a direct flight to Sepahua. (These are small 8-12-passenger planes that usually fly twice weekly. The alternative was to leave Wednesday on a flight bound for Atalaya, then spend some eight to nine hours on a boat on Thursday.) I had planned to head to Kitepámpani on Monday, June 26, again to be delayed due to the very legitimate reason that the family that regularly takes me from Nuevo Mundo, a Matsigenka community on the Urubamba river, to Kitepámpani, came to Sepahua for the birthday of their eldest daughter.
My dissertation has figured little into this blog, and I want to take this opportunity to change that somewhat. My dissertation focuses on the semantics and pragmatics of three second-position clitics in Caquinte -- =sa, =sakanika, and =te -- and their combinations (i.e., =sakanika=te). To a lesser extent it deals with two other second-position clitics -- =tari and =npa -- given that these frequently combine with =te. Despite the grammatical similarities across Kampan languages, the domain of second-position clitics constitutes a notable locus of variation. My dissertation is not comparative or historical in nature, but I am in many ways a historical linguist at heart, and I want to take a few paragraphs here to give some impression of this variation.
The semantics of second-position clitics in Kampan languages are subtle, often consisting of not-at-issue meanings, and as such they are one of the more difficult domains of study in these languages. In practical terms, they are often given very brief descriptions in unexpected sections of dictionaries and/or grammars. Elena Mihas' research on the Perené River dialect of Ashéninka -- in particular her 2015 grammar -- is the single most comprehensive effort to describe these elements. Here, however, I would like to focus on the two Kampan varieties with which Caquinte is presently and has been historically in most contact. I draw on my own field data from Caquinte, as well as early research on the Tambo River dialect of Asháninka conducted by Lee Kindberg (1980:461-466) and corpus-based research on Matsigenka conducted by Lev Michael and myself beginning in 2011 (based on research conducted by him and Christine Beier the same year in Quillabamba, Peru, in conjunction with speaker Haroldo Vargas Pereira). (See Vargas Pereira et al. 2013.) The aim is to give an approximate sense of variation among second-position clitics in these three varieties, based on the data presented in Figure 1. The table is organized based on presumed cognacy, with brief glosses or translations provided for all forms. This is not an exhaustive list of all second-position clitics in each language, nor are the analyses on which glosses and translations are based conclusive. There are almost certainly gaps in the data due to the nature of the sources on which this data is drawn. (In fact, data from Perené Ashéninka, for example, shows some forms cognate to those in Matsigenka.)
 |
| Second-position Clitics in Caquinte, Asháninka, Matsigenka |
We see that only =me, =npa, =ta, and =tari are common to all three varieties. These clitics -- for other reasons not discussed here -- are probably reconstructable to proto-Kampan. Caquinte exhibits eight clitics not attested in the other varieties; Asháninka exhibits seven; and Matsigenka eight. In the context of my own dissertation, we see that Caquinte =sa and =te have cognates in Asháninka but not in Matsigenka, while =sakanika is unique to Caquinte -- the two former are described as markers of "impatience" in Asháninka. Other forms are the clear result of grammaticalization processes that lead to the detection of additional cognates. Caquinte, for example, exhibits three inflectable demonstratives ka, ra, and nta. The proximal demonstrative ka is likely the source of the relativizer =ka. Similarly, a cognate to the Caquinte demonstrative ra is likely the source of the Matsigenka relativizer =rira (-ri being a productive nominalizer in many Kampan varieties). (The Caquinte epistemic modal =ka has a different etymological source, and exhibits reflexes in Nanti (Michael 2008).) Elsewhere cognates are detectable only in morphologically frozen forms, e.g., Caquinte teekera(ta) 'not yet' (cf. dialectal variant teekerota).
Some "long" clitics in some languages are arguably etymologically related to segmentally shorter ones. Caquinte =sakanika is an example of this, although it is not obvious how it could be compositionally analyzed as consisting of =sa, =ka, and =nika. Other possible cases are =nika and =rika; Caquinte =satine and =shiatsi; Matsigenka =kario, =ratyo, =rira, =roro, =tata, and =ukari. I will not speculate about Asháninka etymologies.
I hope to someday carry out a team-based project focused on variation across Kampan languages and dialects. This sort of endeavor has at its core the study of lexical variation, but the morphologically incredibly rich nature of Kampan languages also demands careful comparative study of morphosyntactic differences in order to arrive at the proper understanding of the proto-Kampan verb phrase (second-position clitics included alongside innumerable verbal affixes), as well as issues related to inheritance, borrowing, and the like. Who is interested in fieldwork in the Manú, Urubamba, Apurimac, Ene, Tambo, Pichis, and/or Ucayali driver basins?
References:
Kindberg, Lee. 1980. Diccionario asháninca. SIL.
Michael, Lev. 2008. Nanti Evidential Practice: Language, Knowledge, and Social Action in an Amazonian Society. PhD dissertation, University of Texas at Austin.
Mihas, Elena. 2015. A Grammar of Alto Perené (Arawak). De Gruyter Mouton.
Vargas Pereira, Haroldo and José Vargas Pereira (authors); Lev Michael, Christine Beier, and Zachary O'Hagan (compilers). 2013. Matsigenka Text Corpus (v. June 2013).
Comments
Post a Comment