RINI/TUTEN HISPANIC/ROMANCE LINGUISTICS
On the Romance Future, Grammaticalization,
and the Development of Latin /nVr/
Kenneth J. Wireback
The development of Latin /nVr/ in the Romance future/conditional stems, e.g., VENIRE HABEO > OPtg. verrei, OSp. verré, Ital. verrò ‘I will come’, shows a much greater pan-Romance similarity than the outcomes of /nVr/ in other lexical items, e.g., GENERUM > Ptg. genro, Gal. xenro, Sp. yerno, Ital. genero ‘son-in-law’. This suggests that the reduction of Latin /nVr/ began first in the future/conditional morphology and was subsequently generalized to other word forms. Such an interpretation of the data is in line with the effects of grammaticalization, which suggest that the Latin /nVr/ sequence would be especially susceptible to reduction in a grammaticalization context; thus any sound changes favored by grammaticalization would gain a foothold in the future/conditional stems, and thereby be especially likely to undergo further generalization to other words forms like Latin generum ‘son-in-law’. From this perspective, the reductive effects of grammaticalization favored an assimilatory or metathetic outcome for Latin /nVr/, viz. /nVr/ > /nr/ > /rr/ or /nVr/ > /nr/ > /rn/, since these sound changes depend upon the diminished perceptual salience of the sounds that undergo the change (Hume 2004), e.g., PONERE HABEO > OPtg. porrei, OSp. porné ‘I will place’. In contrast, the requirement that the release burst of an epenthetic stop be sufficiently robust to be perceived, in order to be subsequently incorporated into the underlying form (Recasens 2011), e.g., /nVr/ > /nr/ > [ndr] > /ndr/, suggests that epenthesis (/nVr/ > /nr/ > /ndr/) was impeded by grammaticalization. Consequently, we should expect cases in which /nr/ < /nVr/ outcomes would remain non-epenthetic as /nr/. In fact, the Romance data show this, namely epenthetic forms alongside non-epenthetic outcomes, e.g., TENERE HABEO > Occ. tenrài, tendrài, Cat. tindré, dial. tinré ‘I will have’ (cf. Malkiel 1946: 314 regarding Old Spanish –rn-, -rr-, and –nr- vs. –ndr-).
In this paper I will argue that it is likely that the reduction of Latin /nVr/ began with the grammaticalization of periphrastic PONERE HABEO, TENERE HABEO, AND VENIRE HABEO in Hispano-, Luso- and Gallo-Romance, and that the phonological outcomes favored by grammaticalization were in the majority of cases subsequently generalized to other word forms like GENERUM, TENERUM, VENERIS. In support of this hypothesis I will show that: (1) synthetic Romance future/conditional stems could have developed in all of the Romance areas that would subsequently opt for infinitive + HABEO futures, and that it is not necessary to postulate a Gallo-Romance origin for the synthetic future stems that issued from the grammaticalization of infinitive + HABEO (Valesio 1968; Posner 1996; Company 2006); and (2) there is a consistent relationship in Western Romance between the reduction of Latin /nVr/ in the future/conditional stems and the outcomes of Latin /nVr/ in other adjectival and nominal forms., e.g., metathesis in PONERE HABEO > OSp. porné ‘I will place’ as well as in non-verbal GENERUM > Sp. yerno ‘son-in-law’; lack of epenthesis in PONERE HABEO > Occ. ponrai ‘I will place’ mirrored also in GENERUM > Occ. genre ‘son-in-law’.
Company Company, Concepción, 2006. ‘Tiempos de formación romance II: Los futuros y condicionales’, in Sintaxis histórica de la lengua española, ed. Concepción Company Company (México, D.F.: FCE), pp. 347–418.
Hume, Elizabeth, 2004. ‘The Indeterminacy/Attestation Model of Metathesis’, Language, 80: 203–237.
Malkiel, Yakov, 1946. ‘The Etymology of Hispanic vel(l)ido and melindre’, Language, 22: 284–316.
Posner, Rebecca, 1996. The Romance Languages (Cambridge: C.U.P.).
Recasens, Daniel, 2011. ‘Articulatory Constraints on Stop Insertion in Consonant Clusters’, Linguistics, 49: 1137–1162.
Valesio, Paolo, 1968. ‘The Romance Synthetic Future Pattern and its First Attestations’, Lingua, 20: 113–161.
- Hispanic Linguistics