Tonic Vowels in Vulgar Latin

Linguistics Zone Main Menu

Tonic (ie stressed) vulgar Latin [a]

Classical Latin Ā, Ă > vulgar Latin [a] > Castilian a

PRĀTU > [prato] > prado

CĀRU > [karo] > caro

MĂTRE > [matre] > madre

MĂNU > [mano] > mano

A+I: Because I is palatal there is a tendency to move the A upwards in the vowel triangle to the e position: A+I > [aj >ej] > e

LAĬCU > [lajko] > lego (lay-person, cleric)

(Basque) > [bajka] > vega (water melon; cf Las Vegas)

SAR'TA(G)ĬNE > [sartajne] > sartén (frying pan)

CANTA(V)I > [kantaj] > canté

Intervocalic G was often weak in Classical Latin, and was subsequently lost, producing a diphthong; this was also the case with intervocalic V.

A+U: A+U made its way via a primary wau to the back end of the vowel triangle (o position): A+U > [aw>ow] > o

CAUSA > [kawsa] > cosa

MAURU > [mawro] > moro

AURU > [awro] > oro

TAURU > [tawro] > toro

Secondary wau is the result of velarization and vocalization of L plus consonant; an allophone of the original vowel is produced. Here's the full shebang:

[ał + cons. > aw + cons. > ow + cons.] > o + cons.

BALBU (stammering) > [bałβo > bawβo] > bobo (fool)

'ALT(Ĕ)RU > [ałtro > awtro] > otro

CALCE (heel) > [kałќe > kawќe] > coz (kick)

TALPA > TALPU (theoretical form) > [tałpo > tawpo] > topo

SALTARE > [sałtare > sawtare] > Old Cast. sotar

The movement of the point of articulation to the interdental position at the front of the mouth in CALCE& > &coz is quite substantial. The ł sound is actually the start of the process of fronting. In the case of  SALTARE > O.C. sotar, sal has survived in modern Castilian (salto) to avoid confusion with the homophone which would otherwise have existed (soto - glade). In the case of ALTU > alto and FALSU > falso the l is retained due to frequent use in a liturgical context. Such protection against velarization did not happen of course in Gaul, where ALTU > haut; the h is down to contamination from Germanic languages.

Yod [j] is attracted from the following syllable, in a classic case of metathesis:

[a + cons. + j > aj + cons. > ej + cons.] > e + cons.

'SAPĬA(M) (taste) > [sapja] > sepa (know, but saber is taste; cf also Portuguese)

'CAPĬA(M) (to take) > [kapja] > quepa (to fit)

CAR'RARĬA (cart-track) > [karrarja] > carrera

'BASĬU > [basjo] > beso

No shortage of semantic changes on this occasion either! In the case of 'SAPĬA(M) > sepa it is worth mentioning that in Portuguese, Galician and Leonese the yod is retained. For more extensive parallel comparison, see our section on what happened to tonic V.L. [a] in Romance generally.

Just like yod, wau can also be attracted from the following syllable through metathesis:

[a +  cons. + w > aw + cons. > ow + cons.] > o + cons.

In the first of the examples below, F > H in French and Spanish, before disappearing altogether, but bear in mind that homophones produce resistance; we can also see that haber was (h)aver in Old French, where Old Castilian has an intervocalic bilabial fricative, ove. The H reappeared during Renaissance Latinization.

Initial J was a vowel in  Classical Latin, which became a full consonant in the vulgar vernacular. Also, the concept of regular spelling was not a feature of mediaeval language, such that there are several variations on the sermo vulgaris in square brackets:

'HABI > [aβwi] > O.C. ove > hube

'SAPǓI > [sapwi] > O.C. sope > supe

'JACǓI > [jakwi] > O.C. yogue*

PLACǓI (pleased) > [plakwi] > O.C. plogue

*yacer and placer have regular preterites, (if any at all!)

Tonic vulgar Latin closed e [e]

C.L. Ē, Ĭ, OE > V.L. [e] > Cast. e

Closed [e] was more resistant than [ε] in Romance generally, except in French (bar southern France) where diphthongization occurred ([e > ee > ej > əj > oj > wε > wa]).

In the examples below, the B in CĬBU becomes a fricative, as does the D in FOEDU before disappearing completely:

SĒTA > [seta] > seda

CĬBU > [ќeβo] > cebo

CRĬSTA > [kresta] > cresta

FOEDU > [feðo] > feo

POENA > [pena] > pena

For comparative information, check out the table on vulgar Latin tonic [e] and [o] in Romance.

In hiatus (ie where two vowels are in succession, but belong to separate syllables), [e] is raised to i in Castilian:

VĬA > [βea] > vía

MĔA > MĒA (theoretical form) > [mea] > mía

HABĒ(B)A(M) > [aβea]  > había

DEBĒ(B)A(M) > [deβea] > debía

The Bs and Ms in brackets were extremely weak in vulgar Latin. They were not even uttered in the vulgar Latin of Western Romania, but were much more stable in the East; hence we still find them in modern Arragonese and Italian, albeit as fricatives.

Another popular route from [e] to Castilian i was metaphony, a kind of 'after-sound' similar to the German Umlaut:

VĒNĪ > [βeni] > vine

FĒCĪ > [feќi] > O.C. fize [fidze] > hice NB Imagine the d and z are tied - Microsoft'fault, as is the case for all slight peculiarities present in our transcriptions!

Finally, the [e] can be raised to I through yod development:

'LĬMPĬ(D)U > [lempjo] > limpio

'TĒPĬ(D)U (theoretical form) > [tebjo] > tibio (cf French tiède)

We have already seen Latin B fricate; now we have Latin P becoming occlusive.

Tonic vulgar Latin [i]

C.L. Ī > V.L. [i] > Cast. i

PĪLA > [pila] > pila

VĪNU > [βino] > vino

FĪCU > [fiko] > higo

VĔNĪRE > [βenire] > venire

[i] is very stable in most places:

 eg Latin FĪCU becomes figo (Italian), figu (Sardinian), figo (Portuguese), figa (Catalan), figa/fia (Provençal), fi (Old French) and figue (French).

Tonic vulgar Latin [o]

C.L. Ō, Ŭ > V.L. [o] > Cast. o

CǑLŌRE > [kolore] > color

CǑRŌNA > [korona] > corona

LŬTU > [loto] > lodo

CŬRVU > [korβo] > corvo

[o] can also become u in Castilian through yod-development:'

'CŌ(G)ĬTO > [kojdo] > cuido (Stress is shifted to the i at the final stage.)

'TŬRBĬ(D)U > [torβjo] > turbio

Tonic vulgar Latin [u]

C.L. Ū > V.L. [u] > Cast. u

FŪMU > [fumo] > humo

CRŪDU > [cruðo] > crudo

CŪNA > [kuna] > cuna

MŪRU > [muro] > muro

Tonic [u] is generally resistant in all Romance languages:

eg CRŪDU and FŪMU (Latin) yield the following results:

crudu (Italian), cruu (Sardinian), crud (Romanian and Friulian), cru (Catalan and Portuguese) and cru/crut (Provençal)

fumo (Italian and Portuguese), fumu (Sardinian), fum (Catalan, Provençal, Romanian and Old French), fum (Friulian), and füm (Engadian). The French fumée comes from FUMÁTA.

Tonic vulgar Latin [ε]

Tonic [ε] diphthongizes in Castilian:

C.L. Ĕ, AE > V.L. [ε] > Cast. ie

VĔNTU > [βεnto] > viento

FĔRA > [fεra] > fiera

CAECU > [ќεko] > ciego

QUAERO > [k(w)εro] > quiero (note that initial V.L. k is starting to front here)

In Brazilian and early Portuguese, VĔNTU > vento [vẽntu], but in modern Lisbonese and some areas of Spain one hears [vẽjntu], where Catalan goes for vent [e] and Provençal for ven.

Old Castilian ie > i before elle or s + cons:

ANĔLLU > [anεllo] > aniello > anillo

SĔLLA > [sεlla] > siella > silla

CASTĔLLU > [kastεllo] > castiello > castillo

VĔSPA > [βεspa] > aviespa > avispa

PRĔSSA > [prεssa] > priessa > prisa

'VĔSPĔRA > [βεspera] > viespera > víspera

The latter case is isolated to Castilian. However, FĔSTA > fiesta and GĔNĔSTA > hiniesta

With respect tonic V.L. [ε], let's take a quick look at the reduction of the V.L. diphthong in hiatus before o:

['εo > 'jεo > 'io(>'jo)]

DĔUS > [dεos] > [diaus > dieos] > Dios (Note stress shift.)

MĔU > [mεo] > [miau > mieu] > mío (Here, an early intermediate shift of stress could have gone either way.)

Ĕ(G)O > [εo] >  [yau > ieu] > yo (Again, note stress shift.)

Similarly, JUDAEU > judío

Tonic vulgar Latin [ŏ]

Stressed V.L. [ŏ] diphthongizes in Castilian:

C.L. Ǒ > V.L. [ŏ] > Castilian ue (ua -Hisp. peninsula & Rumanian, uo - this diphthong is not confined to Italian - you'll also find it on the peninsula.)

BǑNU > [bŏno] > bueno

FǑCU > [fŏko] > fuego

RǑTA > [rŏta] > rueda

CǑRPU > [kŏrpo] > cuerpo

Old Castilian ue > e in modern Spanish through dissimilation of the bilabial element in the u (wau); in the first and third cases, it is dissimilated by the labiodental articulation points of initial F, which are very similar to the bilabial:

FRǑNTE > [frŏnte] > fruente > frente

CǑLǑBRA > [kŏlŏβra] > culuebra > culebra (Ptg. cobra: intervocalic l is dropped)

FLǑCCU > [flŏkko] > flueco > fleco (fringe of hair)

Latin velar labials (V and U) produced a lowering effect across the board:

Spanish: ŌVU > ŎVO (theoretical form) > uovo > huevo

Compare: Catalan ou, Provençal ou/uou, Romanian ou, Italian uovo (ovo in dialect), Portuguese ôvo, Old French uef and modern French oeuf.

There are also cases of non-diphthongization:

COMP(Ĕ)RO > [kŏmpro] > compro (exceptionally cuempro in dialect)

CONTRA > [kŏntra] > contra (sometimes cuentra in Old Castilian)

Comprar is not even a radical-changing verb. The lack of diphthongization in this case is probably down to the number of surrounding consonantal elements.

However, FǑNTE > fuente and 'CǑMP(Ŭ)TO > cuento

And finally, proclisis:

DǑM(Ĭ)NU > dŏmn > don

The proclitic term don does not carry its own stress; the stress it might have had is transferred to the word which follows. Speed of articulation is also another factor which may be taken into account when dealing with proclisis, generally.

'CǑM(Ĭ)TE > [kŏmde] > O.C. cuende

                    > [kŏmde] > conde (+name)

'HǑM(Ĭ)NE > [ŏmne] > [uemne > uembre, omne > ombre] > hombre

The H was reinserted at the end of the sixteenth century, but was never uttered. During the Renaissance there was considerable interest in the restoration and recreation of authentic orthography, and although Spanish has not suffered quite so much as French as a result, there is some evidence of the same in Spain.

CǑRĬU > ['kŏrjo] > O.C. & Ptg coiro > cuero

The latter case is just a complete one-off.