Monday, January 31, 2011

'Relational prefixes' and the role of analogy

I've recently come across an interesting example which may demonstrate (if further confirmed) the role played by analogy in creating 'new' morphology, with potential implications for the comparative study of the Macro-Jê and Tupí families (and, possibly, Carib as well). In Araweté (Tupí-Guaraní), the initial consonant in the word for 'ear', reconstructed as *nami by both Lemle (1971:124) and Mello (2002:183), seems to have been reanalyzed as a linking ("relational") prefix: pẽ r-amɨ 'your (pl.) ears', h-amɨ 'his ears', etc. (Solano 2009:105).

Compared to its opposite (a prefix becoming fossilized as part of the root, as documented for Kaingáng; Ribeiro 2004), the analogically-motivated morphological segmentation of a part of the root seems to be much rarer. The only other example I'm aware of, in the "Tucajê" realm, also involves a linking prefix: the word for 'debt' in Karajá--a loan from Portuguese dever--had its initial consonant reanalyzed as a prefix (wa-d-ewe 'my debt', t-ewe 'his debt', etc.). This situation is reminiscent of what occurs with the strong vs. weak distinction in English verbs: although the historical tendency has been towards regularization of strong verbs (e.g. hope), analogy may sporadically lead to the creation of "new" strong verbs (e.g. dived vs. dove, sneaked vs. snuck).

Such examples are useful to demonstrate the relative value of some traditional assumptions in historical linguistics: for instance, although morphological analyzability is generally useful in determining the origin of a given word, it cannot be taken as a full-proof criterion, as demonstrated by Karajá d-ewe 'debt' (from Portuguese dever). And although morphological fossilization seems to be a much more common process (in the case of "relational prefixes" and elsewhere), one has to be prepared for the possibility that analogy may actually work in the opposite direction, extending the reach of fairly non-transparent (even archaic) morphological features.

In more concrete terms, the detection of more or less obvious cases of back-formation such as Araweté r-amɨ 'ear' and Karajá d-ewe 'debt' have interesting implications for the treatment of less straightforward cases of borrowing, such as Karajá dʒ-uwaθa 'poisoned arrow', whose similarity with the word for 'arrow' in Tupí languages (cf. Tupinambá r-uuba-sy 'poisoned arrow', r-uuba 'arrow') seems to suggest a common origin. The fact that they take the linking prefix in both families may at first suggest that both trace back to a common "Proto-Tucajê" form. But, given the attested role of etymology, the possibility of it being a Tupí loan in Karajá cannot be discarded. The fact that the initial consonant is treated as a prefix in Karajá does not necessarily indicate antiquity of the word; it could very well be the result of analogy. [While the word for 'poisoned arrow' in the Tupí example is built upon the independent stem r-uuba 'arrow', the Karajá example is not morphologically analyzable; the basic word for 'arrow' in Karajá is wɨhɨ.]

It would be interesting to know of any additional examples such as the ones described above, involving analogical reanalysis in other Carib, Tupí, and Macro-Jê languages. [For some notes on the comparative status of "relational prefixes", please take a look at my website.]


Lemle, Miriam. 1971. Internal classification of the Tupi-Guarani linguistic family. In: Bendor-Samuel, David (Ed.). Tupi studies I. 1971. p. 107-129. (Summer Institute of Linguistics publications in linguistics and related fields 29). Norman: SIL.

Mello, Antônio Augusto Souza. 2000. Estudo histórico da família lingüística Tupí-Guaraní: aspectos fonológicos e lexicais. Tese de doutorado. Universidade Federal de Santa Catarina.

Ribeiro, Eduardo R. 2004. Prefixos relacionais em Jê e Karajá: um estudo histórico-comparativo. In D'Angelis, Wilmar da R. (org.), LIAMES (Anais do II Encontro Macro-Jê), 4, p. 91-101. Campinas: IEL/Unicamp.

Solano, Eliete de Jesus Bararuá. 2009. Descrição gramatical da língua Araweté. Doutorado, UnB.

No comments: