Academic journal article Bulletin of Hispanic Studies

On Syncope, Metathesis, and the Development of /nVr/ from Latin to Old Spanish

Academic journal article Bulletin of Hispanic Studies

On Syncope, Metathesis, and the Development of /nVr/ from Latin to Old Spanish

Article excerpt

The /nr/ sequence in the spoken Latin of the Iberian peninsula resulted from the syncope of either a posttonic or a pretonic unstressed vowel, and was the precursor to subsequent developments like metathesis to /rn/ or epenthesis of /d/, e.g., generum > /? > Sp. yerno, Cat. gendre 'son-in-law'; tenere habeo > / ?re/ > /ten.?re/ > OSp. tenrre, terné, Cat. tindré 'I will have'. Furthermore, above all with respect to certain future/conditional verb stems, /nr/ was also the precursor to an assimilation process in which the nasal assimilated fully to the following vibrant, e.g., /ten.?re/ > OSp. terré, OPtg. terrei 'I will have'. Schematically, we may represent the aforementioned developments as in (1).

(1) Relative Chronology for Development of Latin /nV. r/

From the initial stage, syncope produced the /nr/ sequence at stage 2, and then one finds either maintenance of the stage 2 /nr/ sequence or subsequent modifi- cation of /nr/ at stage 3, as assimilation, epenthesis, or metathesis. The /rr/, /rn/, and /ndr/ variants at stage 3 are generally assumed to be responses to the relati- vely marked syllable contact or poor sonority profile afforded by post-syncope /nr/ (Penny 2002: 88-90; Harris-Northall 1990: 151, n. 30; Wanner 1989: 437-41; Martínez Gil 2003: 54, n. 1; Ultan 1978: 387-88).

The relative chronology depicted in (1) leaves at least one central question unanswered: Why the /nr/, /rr/ and /rn/ variants at stage 3 were much more prevalent than /ndr/ in the earliest Old Spanish texts. As Malkiel observed, with regard to the future and conditional of poner, tener, and (a)venir, the /nr/, /rn/, and /rr/ variants 'predominate by a wide margin [over /ndr/] in the oldest texts' (1946: 314). Similarly, Malkiel noted that cendrada (< cinerata) 'coarse ashes, plaster of ashes' is less common in Old Spanish than the metathesized variant cernada (1946: 314). Since the relative chronology in (1) places the /rn/, /rr/, and /ndr/ variants at stage 3, as direct reflexes of /nr/, it is not clear why /rn/ and /rr/ should predominate over the epenthetic variant /ndr/. Furthermore, the appeal to an improved syllable contact or sonority profile does not explain the predominance of /rn/ and /rr/ over /ndr/: metathesis, assimilation, and epenthesis all improve the disfavoured /nr/ sequence.

In this article, I investigate in more detail the metathesis that produced /rn/, and propose an inversion process based not only upon a /nr/ > /rn/ shift, wherein the rhotic was a trill, but also a metathesis that applied to [n??] variants of Latin /nV? r/ that retained a vestige of the intertonic vowel and an alveolar tap. These [n??] variants were perceived incorrectly as [?vn] (superscript v represents a svara- bhakti vowel) or [?n] and reanalysed as /rn/. Since the [n??] variants are essentially an intermediate stage between retention of the atonic vowel and syncope, [n??] could have surfaced earlier than the syncopated variant [nr] and become input to metathesis rather than undergoing further vowel weakening to [nr]. This way of viewing the inversion of /n/ and /r/ provides a plausible way to explain why the metathetic /rn/ variants of Latin /nV? r/ were able to predominate over /ndr/ in Old Spanish: Metathesis was able to attain a head start because, unlike consonant epenthesis, it did not have to await the advent of syncope to [nr]. Consequently, metathesis was able to build up early momentum before epenthesis /nr/ > /ndr/ could attain regularity.

In addition to the role of tap realizations of /r/ and variants of Latin /nV? r/ that retained some measure of the intertonic vowel, I propose one additional factor in the early regularization of metathesis in Old Spanish, the reductive effects of grammaticalization on the sequence of phonemes that formed the infinitive + habere constructions. Several types of sound change, including vowel reduction, syncope, and assimilation, can be said to be reductive in nature, a quality that would no doubt be intensified or exacerbated by a grammaticalization context. …

Search by... Author
Show... All Results Primary Sources Peer-reviewed


An unknown error has occurred. Please click the button below to reload the page. If the problem persists, please try again in a little while.