Spanish Phonetics and Phonology The ‘f > h change’ 1

Spanish Phonetics and Phonology 1 Stating the ‘f>h change’ Latin initial /f/ is progressively weakened and eventually lost in standard Castilian in popular words, unless the /f/ is followed by a liquid consonant or a back semivowel ([w]), or in some instances by a front semivowel ([j]). Weakening also sometimes takes place intervocalically. This eventually produces a phoneme split, creating for a time /h/ in addition to /f/, until /h/ is eventually lost. Thus Lat. furnu(m) // > Sp. horno // Lat. f¢ aba // > Sp. haba /a/ Lat. f¢e rru(m) // > Sp. hierro /iero/ - are - // > Sp. sahumar /saumaR/ Lat. s¢ ubfum Lat. defensa // > Sp. dehesa // but Lat. fratre(m )// > Sp. fraile /R/ Lat. f¢ ocu(m) // > Sp. fuego // Lat. f¢e sta // > Sp. fiesta /fiesta/ Lat. f¢ ac¢i le(m) // > Sp. fácil /f/

(before a liquid consonant) (before [w]) (before [j]) (learned borrowing)

2 Phonetic detail We cannot be sure of the phonetic detail of the change. 2.1 A hypothesis A possible general scenario is: [f] > [] > [h] > Ø Penny’s hypothesis (starting from a supposed ?//): Stage 1 (late ‘Vulgar’ Latin of Cantabria) //: [] in the context of following [w] [] elsewhere Stage 2 (Dissimilation of [] before lip-rounded vowels) //: [] in the context of following [w] [h] in the context of following [o], [u] [] elsewhere

Last printed 01/10/04

Spanish Phonetics and Phonology The ‘f > h change’ 2

Stage 3 (Generalization of [h] to appear before all syllabic vowels) //: [] in the context of following [w] [h] in the context of following [i], [e], [a], [o], [u] and [j] [] in the context of following [r], [l] Stage 4 ([] and [] ‘strengthen’ to [f]) /~h/: [f] in the context of following [w], [r], [l] [h] elsewhere Stage 5 [f] and [h] split into phonemes /f/ and /h/ Stage 6 /h/ disappears as a result of [h] > Ø 2.2 The evidence Nebrija’s system of spelling (1492) differentiates between those reflexes of the /f/ which were preserved as [f] and the /f/ which was eventually to be dropped altogether: the former generally being spelt as f and the latter as h. [h] is still preserved in some dialects. Old Castilian texts generally spell all reflexes of /f/ as f (the sporadic use of h increasing with time), which gives us little clue as to the actual phonetic value of f. Penny (1990) has argued that there was a distinction between /f/ and /h/ (or ?//) even when both were spelt with f. 3 The ‘irregularity’ of the change ·

Simple failure of the rule to apply in one of the expected contexts: Lat. foedu(m) // > Sp. feo // Lat. fixu(m) // > Sp. fijo //.

·

Plausible semilearned influence? Lat. f¢e sta // > fiesta // - i u(m) > Sp. febrero /R/ Lat. f¢e br¢ uar¢ Lat. forma // > Sp. horma ‘shoe-last’ /oRma/ / forma ‘form’ /foRma/ Lat. facta /ta/ > Sp. hecha ‘done’ // / fecha ‘date’ //

Variation among speakers: weakening may have been more typical of rural speech (see 4.3). 4 Why did the change occur? 4.1 The substrate hypothesis Latin [f] was modified because Basque learners of Latin found difficulty in pronouncing it. For: · Evidence that Basque in the early Middle Ages had no /f/ phoneme.

Last printed 01/10/04

Spanish Phonetics and Phonology The ‘f > h change’ 3

·

There is a similar f>h change in Gascon, a striking coincidence (both Castile and Gascony bordered the Basque Country).

Against: · /f/ in Latin loanwords in Basque is not adapted to [] or [h] but to a bilabial plosive: Lat. f¢e sta // > Basque besta // Lat. fagu(m) // > Basque pago // · · ·

Later Basque speakers do not seem to have problems in borrowing Castilian words beginning with /f/. Why does /f/ survive in Castilian at all? (see 4.4 for a possible answer) The movement of [f] to [h] is not an unusual change: it is evidenced in a number of southern Italian dialects (Rohlfs 1966:206-7) and is widespread in Andalusia and Latin America. It can easily be seen as an example of articulatory weakening which could have happened quite independently and so needs no appeal to substrate influence at all.

4.2 The structural theory The labial articulation of Lat. /f/ as [] in Castile is parallel to that of Lat. /w/ as [] (cf. the development to [v] in many other Romance languages). (This only explains the first stage of the change.) For: · More consistent with the ‘naturalness’ of the /f/>/h/ change suggested by existence in other languages (see 4.1). Against: • ?// undergoes no parallel subsequent phonetic modification and phoneme split. 4.3 The sociolinguistic theory Words which apparently do not undergo the change are typical of educated, urban speech, whereas those which do are typical of popular, rural speech. Any residual irregularity is due to dialect mixing. For: · The existence of doublet developments such as horma and forma. Against: • Many apparently popular words fail to undergo the change. • Why would urban speakers not have used labiodental [f] more consistently and resisted rural styles of pronunciation? 4.4 Was [f] ‘restored’? If Lat. /f/ came to be pronounced as [] (see 2.1), how is it that [f] exists in modern Spanish? Penny’s (1972) suggestion that adstrate influence may be responsible in the shape of the many influential French and Provençal speakers who came to Castile in the 12th and 13th centuries. These speakers, who settled in the towns, would have had difficulty in articulating

Last printed 01/10/04

Spanish Phonetics and Phonology The ‘f > h change’ 4

the unfamiliar [] sound and might have adapted it to [f], especially in the ‘difficult’ environments before consonants and the back semivowel [w]. It was in the towns as well that words of learned origin would have found first acceptance in the spoken Castilian of educated speakers. Perhaps more significantly, French clerics brought to Castile a new way of reading Latin aloud (the system of ‘litterae’, see Wright 1982:208-20) through which the labiodental [f] may have become familiar. Such circumstances might have favoured the adoption of [f] more generally, and would plausibly explain the many ‘exceptions’ to the f>h rule. Further reading: Catalán de Menéndez Pidal, Diego, 1968. ‘La pronunciación [ihante], por /iffante/, en la Rioja del siglo XI. Anotaciones a una observación dialectológica de un historiador árabe’. Romance Philology, 21, 410-35. Izzo, Herbert J., 1977. ‘Pre-Latin languages and sound changes in Romance: the case of Old Spanish /h-/’, in Michio Peter Hagiwara, Proceedings of the Fifth Linguistic Symposium on Romance Languages (Rowley, MA: Newbury House), pp.227-53 Jungemann, Frederick H., 1955. La teoría del sustrato y los dialectos hispano-romances y gascones (Madrid: Gredos). Malmberg, Bertil, 1971 [1958]). ‘Le passage castillan f > h - perte d’un trait redondant?’, in Phonétique générale et romane (The Hague / Paris: Mouton), pp. 459-63. Martinet, André, 1951-2. ‘The devoicing of the Old Spanish sibilants’, Romance Philology, 5, 133-56. Penny, Ralph J., 1990. ‘Labiodental /f/, aspiration and /h/-dropping in Spanish: the evolving phonemic values of the graphs f and h’, in David Hook and Barry Taylor (eds), Culture in Medieval Spain: Historical and Literary Essays presented to L.P. Harvey (London: King’s College), pp.157-82. Penny, Ralph. J., 1972. ‘The reemergence of /f/ as a phoneme of Castilian’, Zeitschrift für Romanische Philologie, 88, 463-92.

Last printed 01/10/04

Spanish Phonetics and Phonology The ‘f > h change’ 5

5 Appendix: analysis of the sources of Spanish words beginning with f + V Source: Corominas ‘Maintenance’ of /f/: Latin learned (popular doublets in brackets): fanático fábrica fascinar fábula (habla) fatigar facción (hacer) fatuo faceta (haz) fauno facial (haz) fausto fácil (hacer) favor facsímil (hacer) favorable factible (hacer) favorecer factoría (hacer) favorito factótum (hacer) federar factura (hacer) felino facultad (hacer) feliz facundo femenino fagáceo (haya) fémur fámulo

feraz féretro férreo fértil férula férvido (hervir) festinar fétido (heder) feudo fíat (hacer) fiero figura fingir (heñir ‘amasar’) fisco

físico fomento formidable fornicar fortuna fosa fulgor función fundir (hundir) funeral funicular furia fuste fútil futuro

Greek learned faetón fagocito faisán falange falo

fantasía faringe farmacia faro farola

fase fecundo fénix fenómeno fil-

foca fonético fósforo foto-

Arabic faleba falúa (?) fanega faquir (via English or French?) farda fárfara fideo (?) fulano fustán (?)

French fagot falbalá falsete faquín fardo (?) farsa ficha fogoso (?) forraje fusil

Catalan faena faja (?) falda (?) fango farallón feble foja follaje follón fornido forrar (but cf.ahorrar)

Portuguese faca (?) fandango farfullar (?) fayanca

Gallego-Asturiano farruco

Italian facha fachada

Occitan facistol farándula farfante

English faradio folklore fútbol

Possible popular words? (date of first textual attestation in brackets) falso (10th C.) faz (also haz) (10th feria (1100) falta (1220-50) C.) (semiculto?) falto fe (also as he) (1140) fiebre (1220-50) fallecer (1140) fealdad (semiculto?) falla (1140) febrero (semiculto?) fiel fama (mid 10th C.) fecha (also hecha) fiesta (13th C.) familia (1220-50) feligrés (1245) (semiculto?) feo fijo (1256)

fin (1140) (semiculto?) fino (13th C.) firme (1140) fisgar forma (also horma) funda (1335)

Last printed 01/10/04

Spanish Phonetics and Phonology The ‘f > h change’ 6

Onomatopoeic? farra

fofo

?/f/ > /h/ > Ø Latin popular haba hablar hacer hacha hacia hado halcón hallar hambre harina harto hastial hastío haya haz

hazaña hebilla hebra heder helecho hembra heno herir hermoso hervir hez hiel hierro hígado higo

hijo hilo hincar hinojo hito hoja holgar hollar hollín honda hondo hongo horadar hormiga horno

hosco hostigar hoya hoz hozar huir humo hundir huraño hurgar hurto huso

? harapo hato

Last printed 01/10/04