Palatalization in the Romance languages

Palatalization in the Romance languages encompasses a variety of sound changes in Late Latin (alternatively described as Vulgar Latin or Proto-Romance) and in the languages descended from it that caused consonants to gain a palatal or palatalized pronunciation, generally through the influence of an adjacent consonant or vowel. This eventually resulted in the development of a whole series of palatal or postalveolar consonants in most Romance languages, e.g. Italian /ʃ t͡ʃ d͡ʒ t͡s d͡z ɲ ʎ/.

Phonetic basis of palatalization

The term “palatalization” can refer either to a specific phonetic feature, or to a broader class of associated sounds and sound changes. In the strict sense, a consonant is said to be phonetically palatalized when it is pronounced with a secondary articulation where the tongue is raised towards the hard palate, in addition to its primary place of articulation. Various palatalized consonants of this type likely existed at some point in the history of Romance languages; however, as they frequently underwent further changes, their presence often has to be reconstructed rather than being directly attested.

Palatalization can also refer to the replacement of a consonant’s original place of articulation with a palatal one, as in the case of a process that changes [n] (an alveolar or dental nasal) into [ɲ] (a palatal nasal).

More loosely, especially when describing sound changes, the term “palatalization” can refer to other changes in pronunciation that occur as close consequences of palatalization, such as the frequent shift from palatalized or palatal plosive consonants to affricates. For example, original [tj] developed in many Romance languages into the affricate [t͡s]: although the end result is not a palatal consonant, this change occurred in a palatalizing context, and is assumed to result from the former presence of phonetic palatalization. When describing Romance sound changes, “palatalization” is often used loosely to refer to this transition from plosives to affricates, or similar related sound changes.

Conditions leading to palatalization

Palatalization seems to have occurred earliest in the context of consonants followed by the palatal glide [j], which arose from the Latin front vowels ⟨I⟩ or ⟨E⟩ when they were followed by another vowel.

In some contexts, the effects of palatalization are seen in all Romance languages, and can be dated to Late Latin (based on the Romance outcomes and also based on inscriptional evidence). But there are also divergences between languages, both in the conditions where consonants were palatalized, and in the eventual results of palatalization. This implies that palatalization was not a single event. Rather, processes of palatalization occurred multiple times, and in multiple places, across the history of the Romance languages. The outcomes of palatalization depended on the historical stage, the consonants involved, and the languages involved.

In the case of coronal or velar stops (/t k d ɡ/), the initial result of palatalization is often reconstructed as an affricate, such as /t͡ʃ t͡s d͡z d͡ʒ/, which could subsequently develop into a fricative value such as /ʃ s θ z ʒ/. Thus Latin CAELUM[lower-roman 1] 'sky', pronounced with an initial [k], became Italian [ˈt͡ʃɛːlo], Romanian Cer [ˈt͡ʃer], Spanish cielo [ˈθjelo~ˈsjelo], French ciel [sjɛl], Catalan cel [ˈsɛɫ], and Portuguese céu [ˈsɛw].

The original outcomes of palatalization likely continued to be phonetically palatalized even after they had developed into alveolar/postalveolar/etc. consonants. This is shown in French, where all originally palatalized consonants triggered the development of a following glide /j/ in certain circumstances (most visible in the endings -āre, -ātum/ātam). In some cases this /j/ came from a consonant palatalized by an adjoining consonant after the late loss of a separating vowel. For example, MANSIŌNĀTAM > /masʲoˈnata/ > masʲˈnada/ > /masʲˈnʲæðə/ > early Old French maisnieḍe /maisˈniɛðə/ "household". Similarly, MEDIĒTĀTEM > /mejeˈtate/ > /mejˈtade/ > /mejˈtæðe/ > early Old French meitieḍ /mejˈtʲɛθ/ > modern French moitié /mwaˈtje/ "half". In both cases, phonetic palatalization must have remained in primitive Old French at least through the time when unstressed intertonic vowels were lost (?c.8th century), well after the fragmentation of the Romance languages.

Development of Latin postconsonantal /j/

The first step towards palatalization was the development of the Latin front vowels /e/ and /i/ into a palatal approximant /j/ when they were unstressed and followed by a vowel; this occurred regularly in Late Latin.[1]

Subsequent to this change, the consonant preceding /j/ seems to have often been geminated[2] (sustained for a longer duration). For example, Latin FILIUS and VINEA can be reconstructed as developing the pronunciations *[ˈfiːl.ljus] and *[ˈwiːn.nja] respectively.[2] According to one hypothesis, this gemination was a means of resolving an original "unnatural" syllabification *[ˈfiːl.jus] and *[ˈwiːn.ja].[2] Every consonant + /j/ cluster other than /sj/ shows some evidence of lengthening in Romance varieties.[3]

Palatalization and affrication of /tj kj/

All Romance languages exhibit palatalization of original Latin /tj/ and /kj/, which seem to have first developed into affricates, and then undergone a further change into fricatives in some varieties of Romance.[3] Evidence of palatalized pronunciations of /tj/ and /kj/ in Latin is found as early as the 2nd-3rd centuries in the form of interchange between the spellings ⟨ti⟩ and ⟨ci⟩ before a vowel.[4] The development of Latin /t/ into an affricate in the context /tj/ can be dated to the 2nd or 3rd centuries AD;[4] evidence includes inscriptional use of the spellings ⟨tsi⟩ or ⟨tz⟩ in place of ⟨ti⟩.[5]

Affrication of /tj/ seems to have occurred earlier than affrication of /kj/.[4][3][6] The sequence /kj/ was preserved without affrication in some loanwords from Latin into West Germanic (such as Old High German echol and Old Saxon ekil 'steel' from Latin ACIARIUM; Middle High German bracke 'wooden beam' from Latin BRACHIUM, and Old Saxon wikkia 'vetch' from Latin VICIAM).[6] Latin loans into Albanian show a palatal stop /c/ (now spelled ⟨q⟩) as the outcome of /kj/ and of /k/ before front vowels, whereas /tj/ became Albanian /s/ or sometimes /t͡s/ (examples: Albanian faqe 'cheek' from Latin FACIEM 'face'; Albanian kumerq 'toll, duty' from Latin COMMERCIUM 'trade, commerce'; Albanian pus 'well, fountain' from Latin PUTEUM 'well'; Albanian mars 'March' from Latin MARTIUM 'March').[7] The spelling ⟨JUDIGSIUM⟩ for iudicium[4] (Insc. Hisp. Christ. 108), which can be dated to the sixth century, may be the earliest direct evidence for affrication of original /kj/.[8]

Outcomes of /tj/ versus /kj/

In Tuscan, Corsican, and some Rhaeto-Romance languages, the outcome of Latin /tj/ is a more anterior (alveolar) affricate than the outcome of /kj/, whereas in other varieties of Romance, /tj/ and /kj/ developed into affricates or fricatives sharing the same place of articulation.[4]

In Romanian, [t͡s] is the general outcome of both /tj/ and /kj/ (except for before a stressed syllable, where /tj/ became [t͡ʃ]), whereas [t͡ʃ] is the outcome of /k/ before /i/ or /e/. Balkan Romance can be reconstructed as initially merging /tj/ and /kj/ without affrication as /(t)tj/, and subsequently undergoing a single wave of palatalization that occurred late enough to affect secondary /tj/ derived from diphthongization of Latin /e/, as in Latin TERRAM > *[ˈtjɛrra] > Romanian țară 'country'.[9]

In Sardinia[10] and Southern Italy,[11] the original outcomes of both /tj/ and /kj/ after palatalization can be reconstructed as [t͡s] or [tt͡s].

In many Western Romance languages, Latin intervocalic single stops came to be voiced, but intervocalic /kj/ developed into a voiceless consonant. This implies that /kj/ underwent gemination in Western Romance.[12] In contrast, intervocalic /tj/ can have a voiced outcome,[13] which is assumed to result from the earlier palatalization of /tj/ compared to /kj/.[3] However, there are a number of cases where intervocalic /tj/ developed instead into a voiceless consonant in Western Romance, merging with the outcome of /ttj/[13] or /kj/. There are several proposed explanations for the variable outcome of intervocalic /tj/ in Western Romance. One hypothesis is that /tj/ was lengthened to /ttj/ in certain specific words;[5] for example, that Catalan plaça derives from *plattea,[12] from Latin PLATEAM. Another hypothesis is that the voiceless outcome results from confusion between Latin /tj/ and /kj/[14] (perhaps at a time when [tʲ] or [c] was a potential realization of either sequence).

There are competing hypotheses about the origin in Western Romance languages of [t͡s] as the outcome of /kj/ (and also of /k/ before /i/ or /e/, for which see below).[4] According to one view, the Western Romance outcomes result from Latin /kj/ initially developing into the affricate [t͡ʃ][4] (or [tt͡ʃ][13]),[15] followed by depalatalization of [(t)t͡ʃ] to [t͡s],[4][16] but the existence of [t͡ʃ] as an intermediate step in this context is disputed.[3] According to an alternative hypothesis, the original velar stop in /kj/ became a phonetically palatal stop [c], which was reidentified as /tj/ and subsequently affricated to [t͡s].[4]

Outcomes of Latin -tj- and -kj- in intervocalic position[17]
BranchLanguage-tj--kj-
SardinianCampidanese~Central Sardinian~Logudoreset͡ːs~θː~tː
Western RomanceWest/North Friuliant͡ʃ
East Friulians
Fassant͡st͡ʃ
Comelicanð/θθ
Livinallongheset͡st͡ʃ
Surselvan, Sutselvan, Surmiran, Engadiniant͡st͡ʃ
Venetianss/θ
Ligurians (t͡s)
Lombardss/ʃ
Picardʃ
Frenchjzs
Franco-Provençalz (ʒ, θ)s (ʃ, θ)
Auvergnat, Occitanzs
Catalanz/ð[lower-roman 2]s
Spanishθ[lower-roman 3]
Portuguesezs
OtherRomaniant͡s (t͡ʃ)
Vegliotes
Calabreset͡ːs
Tuscan and Corsicant͡ːst͡ːʃ

/dj ɡj/

The sequences /dj/ and /ɡj/ could both merge with /j/ as the result of lenition.[2] (Note that between vowels, Latin /j/ was in fact normally a geminate consonant [jj], as in MAIUS [ˈmajjus], although many of the Romance outcomes imply development of intervocalic /j/ into a single consonant.[2]) In the case of /dj/, it seems to have been possible either for it to be reduced to [j] early on, or for it to survive long enough to develop parallel to /tj/.[2] The outcome could depend on the position in a word.

  • Romanian fairly consistently has /z/ < /d͡z/ from palatalized /d/.
  • In Standard Italian, /ɡj/ becomes /dd͡ʒ/ and /dj/ becomes either [(d)d͡ʒ] or [(d)d͡z] (geminate between vowels, single otherwise). The outcome of /dj/ is always [d͡ʒ] at the start of a word, and always [d͡z] after a consonant[18] (with the exception of [verˈɡoɲɲa] 'shame'[19] from Latin VERECUNDIAM): examples include DIURNUM > giorno [ˈd͡ʒorno] and HORDEUM > orzo [ˈɔrd͡zo]. Between vowels, /dj/ usually results in [dd͡ʒ], but sometimes [dd͡z].[18] Since [(d)d͡ʒ] is also the regular outcome in standard Italian of Latin initial and intervocalic /j/, these results seem consistent with an original merger of /ɡj/ and (initial or intervocalic) /dj/ with /j/, followed by fortition of /j/ to [(d)d͡ʒ].[20] There is no clear explanation for why intervocalic /dj/ sporadically resulted in [(d)d͡z].[21]
  • In Sardinian, /dj/ seems to have merged with /j/ in all contexts.[10]
  • In French, the outcomes of /dj/ and /ɡj/ appear consistent with early simplification to /j/ in all positions, followed by fortition of /j/ in word-initial position or after /r/, yielding modern French /ʒ/, as in DIURNUM > jour [ʒuʀ] and HORDEUM > orge [ɔʀʒ].[22] Note however that original /rj/ (see below) shows a different outcome in words such as aire < AREAM and cuir < CORIUM.[23] The sequence /ndj/ developed to /ɲ/ (also the usual outcome of /nj/, for which see below), as in BURGUNDIAM > Bourgogne and VERECUNDIAM > vergogne.[23]
  • In Spanish, original /dj ɡj j/ merged between vowels; the original outcome was likely a geminate palatal consonant [ʝʝ], which was subsequently simplified along with other geminates and became the consonant spelled ⟨y⟩ in modern Spanish.[24] This was lost after a front vowel in early Old Spanish, as in SEDEAM > sea, CORRIGIAM > correa, and PEIOREM > peor.[25] After a consonant, /dj/ developed into the Old Spanish voiceless affricate [t͡s], as in HORDEOLUM > orçuelo, VIRDIAM > berça, VERECUNDIAM > vergüença (but alternatively vergüeña).[26] Latin initial /j/ became /j/ (phonetically [ɟ͡ʝ]) except before /o/, /u/; nearby Gascon is similar.[27] This suggests that Spanish and Gascon (originating from isolated districts behind the western Pyrenees) were relic areas unaffected by a fortition of /j/ > /(d)d͡ʒ/ seen in other Romance languages.
Outcomes of Latin /dj ɡj j/ in intervocalic and initial position[28]
BranchLanguageintervocalicword-initial
-ɡj--j--dj-dj-j-
SardinianCampidanese∅d͡ʒ
Central Sardinian∅j
Logudorese∅jj/d͡ʒ/d͡z
Western RomanceWest/North Friulianj/∅j/zɟ/j/d͡ʒj/(d)ʒ/d͡z
East Friulianj/∅j/zj/ʒ/z
Fassanj/∅ʒzʒ
Comelicanjd/ð
Livinallonghesejj/zʲd͡ʒʒ
Surselvanɟɟ/jd͡zɟ
Sutselvanɟ(ʒ)zɟɟ/ʒ
Surmirand͡z
Engadinianɟ/jd͡zɟ/j
Venetianzz/ðd͡ʒ/zz/ð
Liguriand͡zz(d͡z)d͡zd͡ʒz
Lombardzʒ/zz(d)ʒ
Picardjʒ
Frenchjʒ
Franco-Provençaljj/d͡zd͡z
Auvergnatd͡z
Occitanʒ
Catalan(d)ʒ
Spanishj/∅jj/x
Portugueseʒ
OtherRomanianj/∅jzʒ
Veglioted͡zj/∅d͡z
Calabresej
Tuscandːʒdːʒ, dːzd͡ʒ
Corsicanɟɟ/jɟ/dːzɟɟ/d͡ʒ

Other /Cj/ sequences

The palatalization of dental and velar stops (/t d k ɡ/) before /j/ seems to have been followed by a second wave of palatalization that affected other consonants followed by /j/.[2] The relatively later date of palatalization in this context is supported by differences in the development of preceding vowel sounds in Old Occitan: the sound change of metaphony targeted open-mid vowels that were followed by /j/, and its effects are seen in fuelha from Latin FOLIA but not in forsa from Latin FORTIAM.[2] This implies that at the time Old Occitan metaphony took place, /j/ had been absorbed into the preceding consonant in the case of /tj/, but not yet in the case of /lj/.

/nj/

Central Italian,[29] Southern Italian,[11] and Western Romance[15] languages all show development of the sequence /nj/ into the palatal nasal consonant /ɲ/. In Central and Southern Italian, this is pronounced more specifically as the geminate consonant [ɲɲ], and the evolution of preceding vowel sounds in Western Romance implies that /ɲ/ was originally a geminate consonant in these languages as well.[15] The gemination can be interpreted as resulting from an intermediate stage such as [n.nj].[30]

In Balkan Romance, both /nj/ and /ni/ underwent palatalization.[9] In Romanian, the resulting /ɲ/ came to be denasalized to /j/ or deleted, as in CALCANEUM > Romanian călcâi 'heel' or VINEAM > Romanian vie 'vineyard'[31] (but /n/ was analogically restored in ani 'years', which replaced ai, the regular descendant of Latin ANNI).[9] The palatal nasal /ɲ/ is retained in Aromanian and in the region of Banat north of the Danube.[9]

In French, a few[32] words show an alternative outcome with the fricative [ʒ], which corresponds to an original [d͡ʒ][33] in Old French. (Notably, this is identical to the regular outcome of /mj/,[34] for which see below.) Examples include EXTRANEUM > French étrange 'strange',[33] LINEUM 'linen' > French linge [lɛ̃ʒ],[35][33] and LANEUM > French lange.[36] This outcome may represent cases where /j/ did not palatalize the preceding /n/, but was strengthened into an affricate instead;[37] alternatively, it has been explained as resulting from the affrication of a palatalized nasal[35][36] (via a sequence of changes such as /nj/ > /nʲ͜dʲ/[36] > /nd͡ʒ/). It has also been suggested that the words in question are 'learned',[38] that is, borrowed from Latin early[34] and subjected to the vernacular sound-changes /i e/ > /j/ and /j/ > /d͡ʒ/.[39] (The sequence /mnj/ regularly developed to /◌̃ʒ/ in French,[40] like /mj/. Compare the development of /mn/ to /mm/, shown by words like SOMNIC(U)LUM > French sommeil.[40])

In Sardinian, original /nj/ developed into a cluster of a nasal and voiced affricate (e.g. VINEAM > Nuorese [ˈbind͡za],[10] Campidanese [ˈbind͡ʒa],[41] Logudorese [ˈbind͡za][41][36]). A similar outcome is found in some southern varieties of Corsican (e.g. VINEAM > [ˈvinɟa][29]). As in French,[36] the use of nasal + affricate clusters such as [nd͡z] in Sardinian has been interpreted either as the result of reinforcement of syllable-initial /j/ in the cluster /nj/ without palatalization of the nasal,[10][35] or as the result of palatalization of /nj/ followed by reinforcement of the resulting palatalized consonant.[36]

/lj/

The sequence /lj/ became the palatal consonant /ʎ/ (with gemination) throughout Western Romance as well as in Southern and Central Italy.[42]

In Ibero-Romance, /ʎ/ survived as such in the east (Aragonese) and west (Galician-Portuguese). In the centre it yielded either /j/ (Astur-Leonese) or /x/ (Spanish, reflecting an original /ʒ/). The Old Spanish /ʒ/ seems most likely to have arisen from an older stage /j/, although a direct development from /ʎ/ has also been suggested.[43]

In Balkan Romance, palatalization of /lj/, as of other clusters in /j/, appears to have taken place at a relatively late date (after the degemination of -LL- and the diphthongization of [ɛ] to [jɛ]), and /l/ was also palatalized before /i/.[9] The resulting /ʎ/ was retained in Aromanian, but delateralized to /j/ in Romanian.[9] The following examples show the Romanian outcome: FOLIA > *[ˈfɔʎa] > Romanian foaie ‘leaf’, LEPOREM > *[ˈljepure] > *[ˈʎepure] > Romanian iepure 'rabbit', and GALLINAM > *[ɡaˈʎina] > Romanian găină 'hen'.[9]

In some Sardinian varieties, the ultimate outcome of /lj/ is a geminated voiced affricate, as in Logudorese [ˈfɔdd͡za][10] and Campidanese [ˈfɔdd͡ʒa][10] from Latin FOLIA. These can be interpreted as resulting either from palatalization of /l/ followed by affrication of the resulting palatal lateral,[36] or from fortition of a syllable-initial /j/ (as after /n/) followed by assimilation of the preceding /l/ (e.g. *[ˈfɔl.ja] > *[ˈfɔld͡za] > [ˈfɔdd͡za][10]). The dialect of Cagliari has [ll], which probably developed by depalatalization of former [ʎʎ].[10]

/rj/

In Western Romance, /rj/ universally developed via [rʲ] to [i̯r][15] (which can also be written [jr] and interpreted as a case of metathesis[35]). French however shows the cluster /ʀʒ/ as an alternative outcome in some words, such as CEREUM > cierge and BURRIONEM > bourgeon.[44]

Italo-Romance languages show various outcomes including loss of the /r/, loss of the /j/ element, and gemination to /rr/.[35]

In Balkan Romance, [rʲ] seems to have developed variously into [rj], [r], and [j].[9]

Pronunciation of original /rj/ as a cluster is found in Sardinian, with fortition of /j/ in some varieties (e.g. CORIUM > Nuorese Sardinian [ˈkorju] 'leather', Logudorese [ˈkord͡zu], Campidanese [ˈkord͡ʒu])[10] and in some southern Corsican varieties (e.g. AREAM > [ˈarɟa]).[29]

/sj/

Intervocalic /sj/ (including /nsj/, since Latin /n/ was generally lost early on before /s/) shows the following outcomes:

  • Portuguese has [jʒ], as in CASEUM 'cheese' > [ˈkejʒu][45] or BASIUM > beijo 'kiss'.[46]
  • In Spanish, the outcome can be traced back to [jz]. The [j] combined with a preceding vowel, forming diphthongs that ended up being simplified, and the [z] was eventually devoiced to [s]. Examples of this development are queso 'cheese' < *[ˈkejzo] < *[ˈkajzo] < CASEUM and beso 'kiss' < *[ˈbejzo] < *[ˈbajzo] < BASIUM.[47]
  • In French, /sj/ developed to [jz], as in MANSIONEM > maison 'house',[48] probably via the intermediate stage of a palatalized sibilant such as [zʲ]. The [j] combined with a preceding vowel, forming diphthongs that ended up being simplified.
  • In Tuscan Italian, intervocalic /sj/ developed to [ʃ] or sometimes to voiced [ʒ];[49] these came to be reinterpreted as the phonemes /t͡ʃ/ and /d͡ʒ/ after these developed intervocalic allophones [ʃ] and [ʒ] in Tuscan.[50] In Old Tuscan texts, intervocalic /ʃ/ and /ʒ/ were still phonemically distinct from /t͡ʃ/ and /d͡ʒ/, and spellings like ⟨sc(i)⟩ and ⟨sg(i)⟩ are found for /ʃ/ and /ʒ/ respectively.[51] After the lenition of intervocalic affricates (which seems to be attested by confusion in spelling since the second half of the 13th century for /d͡ʒ/, and since the early 15th century for /t͡ʃ/), the usual spellings of intervocalic [ʃ] and [ʒ] (phonemically /t͡ʃ/ and /d͡ʒ/) in Tuscan texts came to be ⟨c(i)⟩ and ⟨g(i)⟩, as in ⟨bacio⟩ and ⟨pertugio⟩.[50] In Standard Italian, phonetic affricates [t͡ʃ] and [d͡ʒ] are found, likely as the result of spelling-pronunciation by speakers outside of Tuscany.[52]
  • In Romanian, /sj/ became /ʃ/, as in CASEUM > caș 'cheese'.[9] This is also found as the outcome of /s/ before the vowel /i/, as in RESINAM > rășină 'resin'.[9]
  • In Sardinia and the south of Italy, /sj/ developed to /s/[45][53] (sometimes voiced to [z]). Examples of this outcome are Nuorese Sardinian [ˈkazu] and Neapolitan caso 'cheese' from Latin CASEUM.[53]

Geminate /ssj/ could develop into [ʃʃ] (as in Old Florentine grascia from *CRASSIAM[29]); this outcome is found in some varieties that show a non-palatal outcome for intervocalic /sj/, such as Neapolitan (with avasciare 'to lower' < *BASSIARE[11]). Per Recasens, such cases of asymmetrical development may be the result of phonetic factors that make palatalization less favored for voiced compared to voiceless consonants.[54] The sequence /rsj/ could have the same outcome, as in Tuscan [roveʃˈʃaːre] from Latin *REVERSIARE[50] (compare the development of -RS- to /ss/ in DORSUM > Italian dosso).

Labial + /j/

The sequences /pj bj vj mj/, which feature a labial consonant followed by /j/, show various outcomes in the Romance languages.[55] In Vulgar Latin, /bj vj/ merged in intervocalic position as [βj];[56] it seems [β] in this context was sometimes lost or delabialized early on, and as a result, Latin /bj vj/ > [βj] sporadically has the same outcome as Latin intervocalic /j dj ɡj/. (Examples of this outcome in French include HABEO > ai and DEBEO > dois.[57] Examples in Spanish include FOVEAM > Old Spanish foya and HABEAM > haya.[58]) In a larger set of words, the labial consonant was initially retained, but underwent diverse developments in combination with the following /j/ in different Romance languages.

Gemination

In Italian, intervocalic /pj bj vj mj/ show gemination of the labial consonant, resulting in /ppj bbj bbj mmj/ respectively.[56] Examples include SAPIAT > [ˈsappja],[35] RABIAM > rabbia[56] [ˈrabbja],[35] HABEAT > abbia, CAVEAM > gabbia, VINDEMIAM > vendemmia.[56]

Western Romance data shows inconsistent application of gemination in intervocalic labial + /j/ clusters;[15] some forms such as Spanish jibia[15] 'cuttlefish' from Latin SĒPIAM show the effects of intervocalic lenition on the labial consonant, implying a lack of gemination. (Penny considers it likely that the form jibia is Mozarabic in origin rather than a native Castilian development.[59])

Metathesis

Portuguese exhibits what is traditionally described as ‘metathesis’ of labial + /j/ sequences: the labial consonant is retained but the /j/ is moved before it. Examples include APIUM > aipo 'celery', RABIAM > raiva 'anger, rage', RUBEUM > ruivo 'red-haired', and NOVIUM > noivo 'fiancé'.[60] It has been argued that the labial consonant and palatal glide did not switch positions in a single abrupt step, but underwent the following series of sound changes:

  1. First, labial + /j/ sequences coalesced into the palatalized labial consonants [mʲ pʲ bʲ vʲ]. Spellings such as ⟨mh⟩, ⟨vh⟩ and ⟨bh⟩ may attest to the original development of palatalized consonants in this context[61] (compare the still-current use in Portuguese of ⟨nh⟩ and ⟨lh⟩ as spellings for palatal /ɲ/ and /ʎ/).
  2. Next, an epenthetic glide [j] developed between a vowel and a following palatalized labial consonant.
  3. Finally, palatalized labial consonants were depalatalized, becoming plain labials preceded by a (now phonemically distinct) palatal glide.

It appears that these changes occurred between Old and Medieval Portuguese, at a later date than the palatalization and ‘metathesis’ of /sj/, /zj/ and /rj/ in Hispano-Romance:[61] metathesis of the preceding sequences is found regularly in both Spanish and Portuguese, and was followed by a shift from [aj] to [ej] that can be seen in Portuguese queijo, eira, queixar, whereas metathesis of labial + /j/ occurs regularly in Portuguese but not in Spanish, and affected Portuguese words show unshifted /aj/.[60] The Portuguese metathesis of labial + /j/ sequences occurred late enough to affect some cases of secondary /j/ that developed after lenition of a following intervocalic consonant (as in LIMPIDUM > *[ˈlim.pjo] > limpho 'clean' and COMEDO > *[ˈko.mjo] > coimo 'I eat').[61] In cases where a palatalized consonant came after another consonant or after the vowel /i/ (e.g. modern Portuguese limpo 'clean'), the original /j/ may be attested only indirectly in the modern language by its effect of raising a preceding vowel (metaphony).

In Spanish, Latin labial + [j] sequences did not systematically undergo metathesis; the general outcome is simply a labial consonant followed by [j].[lower-roman 4] This is shown by the following examples: APIUM > apio 'celery', RABIAM > rabia 'anger, rage'; RUBEUM > rubio 'blond', NOVIUM > novio 'boyfriend'.[60] However, metathesis of original [pj] to [jp] occurred in forms of two Spanish verbs, saber 'to know' and caber 'to fit': the effects of this metathesis are seen in forms like sepa (from SAPIAT) and quepo (from CAPIO).[62] Wireback argues that in Spanish, unlike in Portuguese, there was an abrupt inversion from /pj/ to /jp/ in these verb forms as a result of morphological analogy with vowel + /j/ sequences found in the inflectional paradigms of other verbs.[63] These two verbs also show metathesis of /pw/ to /wp/ in their preterite stems. For example, saber in Old Spanish had the preterite form sope 'I knew' (from *[ˈsawpi] < *[ˈsapwi] < SAPUI) (/o/ here was later replaced analogically by /u/, yielding modern Spanish supe).[64] Even though the cognate Portuguese verbs saber and caber also show metathesis of labial + /j/ and labial + /w/ sequences (in forms such as saiba, caibo, soube, coube), it is thought that metathesis occurred in Spanish and Portuguese at different times, rather than occurring once in their common ancestor.

Proto-Romanian shows the development of a diphthongal offglide after a stressed vowel followed by an original sequence of labial consonant + palatal glide, as illustrated by *scupio > Romanian scuip, HABEAT > Romanian aibă, and DIFFAMIAM > Romanian defaimă.[65] The glide remained after an unstressed syllable, as in APPROPRIARE > Romanian apropia.[65]

Glide strengthening

In various Romance languages, labial consonants in a palatalizing context show outcomes that seem to result from strengthening of a glide to a palatal obstruent (sometimes accompanied by complete loss of labiality):

  • Some Balkan Romance languages, after the split of Proto-Romanian, show the development of pronunciations like [(p)kʲ~(p)tʲ~(p)t͡ʃ], [(b)ɡʲ~(b)dʲ~(b)d͡ʒ], and [mnʲ~nʲ] from labial consonants followed by /j/ or /i/, as in /koˈpil/ [kopˈkʲil] 'child'.[66] These seem to have arisen from palatalized labials such as [pʲ], [bʲ], [mʲ] by 'consonantification' of the offglide.[66]
  • Old Provençal shows [ˈrobd͡ʒe] from Latin RUBEUM and [ˈsapt͡ʃa] from Latin SAPIAT, and the Lombard dialect of Borno shows [ˈbd͡ʒulk] from *bjulk ‘yokel’.[35]
  • Old French shows /t͡ʃ/ as the usual outcome of /pj/; /d͡ʒ/ as the outcome of /bj βj/; and /nd͡ʒ/ as the outcome of /mj mbj mnj/. These correspond to modern French /ʃ/, /ʒ/, and /◌̃ʒ/ respectively. The following examples demonstrate these outcomes: SĒPIAM 'cuttlefish' > seiche,[67] SAPIAM > sache,[68] RUBEUM 'red' > rouge,[69] CAVEAM > cage,[69] SALVIAM > sauge,[57] SERVIENTEM > sergent,[57] SĪMIUM 'monkey' > singe,[69][41] CAMBIARE > changer,[56] SOMNIUM > songe.[40] The Old French pronunciations are likely derived from simplification of labial-affricate sequences[35][15] such as [pt͡ʃ bd͡ʒ md͡ʒ],[70] or [vd͡ʒ].[69] These may have developed from palatalized labial consonants by means of offglide consonantization (as in Balkan Romance);[69] e.g. [pʲ] > [pt͡ʃ] > [t͡ʃ]. A competing explanation of the French outcomes is that Latin /pj bj mj/ remained clusters, and then the postconsonantal /j/ underwent fortition (with the resulting affricate being assimilated in voice to the preceding consonant).[70]
  • In Neapolitan (in southern Italy), the ultimate outcome of /pj/ and /bj/ are geminate affricates as in SĒPIAM > seccia and RABIAM > arraggia 'rage'.[11] These may have developed from labial-affricate sequences [pt͡ʃ] [bd͡ʒ]; an alternative explanation is that geminated palatalized labials [ppʲ] [bbʲ] were reinterpreted as palatal consonants due to acoustic similarity.[11]

/k ɡ/ + front vowel

The Latin velar stops /k ɡ/ developed palatalized outcomes before the front vowels /i e ɛ/ in all Romance varieties except certain dialects of Sardinian and Dalmatian.[71] Palatalization in this context can be dated to about the fifth century AD,[41] although it is possible that it occurred independently and at a later date in eastern Romance compared to western Romance.[72][73] In Romanian, unlike most Romance languages, palatalization occurred after the loss of labialization in sequences of [kw] or [ɡw] + front vowel; thus Romanian shows affricates in words such as sânge and acel < SANGUEM and *ECCUM ILLUM.[74]

The Ragusan dialect of Dalmatian shows no palatalization of /k ɡ/ before any vowel.[75] The Vegliote dialect of Dalmatian shows palatalization of /k/ to /t͡ʃ/ before /i/, but this is argued to be an innovation that arose internally in this dialect, not a trait that it shares by common inheritance with other palatalizing Romance varieties.[75] The Vegliote Dalmatian palatalization also occurred before the glide [j] found as the first element of the diphthongs /i̯e i̯a/, as in [munˈt͡ʃal] 'hill', which is assumed to have developed by means of *[munˈkjel] from Latin MONTICELLUM.[75]

/ɡ/

The palatalization of voiced /ɡ/ before a front vowel may have begun earlier than the palatalization of voiceless /k/.[76] Epigraphic evidence indicates that in Latin of the late Empire onwards, intervocalic /ɡ/ may have already been lost in some words where it occurred between non-back vowels,[77] for example in viginti, frigidus, digitus[77] or legit, sagitta.[2]

In most Romance languages, palatalization of /ɡ/ before a front vowel resulted in the same outcome as original /j/ and the sequences /dj/ and /ɡj/.[71] Exceptions to this merger include Romanian and some Rhaeto-Romance varieties.[71]

The palatalization of the sequence /nɡ/ by a following front vowel has resulted in many cases in a palatal nasal [ɲ(ɲ)] (the same outcome as original /nj/); alternatively, /nɡ/ could develop in this environment into a nasal followed by an affricate or fricative, depending on the language (or sometimes, on the word). The regular outcome of a nasal + [ɡ] + a front vowel is [ɲɲ] across nearly all of southern Italy.[78] In contrast, dialects of northern Italy show [nd͡z] or [nz], which probably developed from earlier [nd͡ʒ].[78] In Tuscany, both [ɲɲ] and [nd͡ʒ] are found. Their original distribution seems to have been based on geography, with [ɲɲ] in eastern Tuscany (and in Old Florentine) and [nd͡ʒ] in the west.[78] However, modern Florentine has [nd͡ʒ], and Old Florentine shows a mixture of forms such as piange alongside piagne 'he cries' and spegnere alongside spengere 'to extinguish'[78] (from Latin PLANGERE and EXPINGERE[79]). The reasons for the displacement of [ɲɲ] by [nd͡ʒ] in Florentine are unclear, but it may have been due to influence from northern Italian and from the regions of Tuscany where [nd͡ʒ] was the regular outcome.[78] Standard Italian, like modern Florentine, generally has [nd͡ʒ] (with the exception of spegnere).[78] In Spanish, /nɡ/ + front vowel shows three possible outcomes: Old Spanish /nd͡z/ (modern Spanish /nθ/), as in GINGIVAM > Old Spanish enzia > modern Spanish encía 'gum'; /ɲ/, as in RINGERE > reñir 'to scold'; and /n/, as in QUINGENTOS >quinientos.[80] The development to [ɲ] seems to be typical in verbs.[81]

There are relatively few examples of the outcome of [lɡ] before a front vowel in Italian[82] and Spanish.[80] The sequence -LIG- in the Latin verb forms COLLIGIT and *EXELIGIT developed the same way as original [lj]. This has been cited as evidence that [lɡ] developed to [(ʎ)ʎ] before a front vowel, based on the assumption that -(L)LIG- here underwent syncope to [lɡ]; however, it is possible that these outcomes instead reflect the aforementioned early loss of intervocalic /ɡ/ between non-back vowels, followed by a change of the prior vowel into a glide (yielding [lj]). In the case of Italian, Rohlfs[79] (1966) cites cogliere from COLLIGERE and scegliere from *EXELIGERE as evidence that a palatal lateral was the usual outcome in Tuscan of [lɡ] before a front vowel (words such as algere, emulgere, fulgere, indulgere are disregarded as non-inherited forms).[83] Likewise, Maiden[79] (2013) derives Italian [ˈkɔʎʎe] 'he plucks' and [ˈʃeʎʎe] 'he chooses' from *[ˈkɔlɡe] and *[ekˈselɡe].[82] On the other hand, Boyd-Bowman (1980) traces Italian cogliere 'collect' to Latin COLLI(G)ERE and views it as showing the same development as in ALLIUM > aglio.[84] As for Spanish, although Hanssen (1913) derives Spanish coges from a syncopated form *COLGIS,[85] Penny (2002) prefers the derivation COLLIGIS > *[ˈkɔllees] > *[ˈkɔlljes] > *[ˈkɔʎes] > coges 'you grasp',[86] and Tuttle (1912) argues that coge does not descend from *COLGET, but instead goes back to a form like *colliet that developed by elision of intervocalic [ɡ].[87] Malkiel (1982) notes the scarcity of examples of the outcome of [lɡ] + front vowel in Old Spanish and considers there to only be a single indirect example of its outcome, the modern Asturian verb esmucir, tentatively assumed to descend from EXMULGERE via an intermediate stage of *esmulzir.[88]

The sequence /rɡ/ before a front vowel usually developed into [rd͡z] in Old Spanish.[80] In Tuscany, the outcome is [rdʒ] (except for Old Tuscan ariento, an alternative form of argento 'silver'),[82] while [rj] is found in southern Italy.[82]

/k/

Palatalization of the Latin voiceless velar plosive /k/ before /i, e, ɛ/ appears to have initially resulted in an affricate, either /t͡ʃ/ or /t͡s/.[72] The value /t͡ʃ/ is found in Italian and Romanian; /t͡s/ or a derivative of it is found in many Western Romance languages, as well as in the Eastern Romance language of Aromanian. It is often suggested that /t͡ʃ/ was the original result in all languages, with /t͡ʃ/ > /t͡s/ a later innovation in the Western Romance languages.[89] A suggestive fact is that the Mozarabic language in al-Andalus (modern southern Spain) had /t͡ʃ/ as the outcome despite being in the "Western Romance" area and geographically disconnected from the remaining /t͡ʃ/ areas; this can be explained by the hypothesis of a sound change from /t͡ʃ/ to /t͡s/ that had not reached the area where Mozarabic was spoken. (Northern French dialects, such as Norman and Picard, also had /t͡ʃ/, but this may be a secondary development, i.e. due to a later sound change /t͡s/ > /t͡ʃ/.)

In Western Romance, Latin intervocalic /k/ before a front vowel was affected by both palatalization and intervocalic voicing,[15] and so generally had a distinct outcome from word-initial or post-consonantal /ki ke kɛ/. (In Spanish, this is disguised by the eventual devoicing of Old Spanish ⟨z⟩ [d͡z], which resulted in the replacement in modern Spanish spelling of Old Spanish -⟨ze⟩- -⟨zi⟩- with -⟨ce⟩- -⟨ci⟩-.)

Some attested outcomes of /k ɡ/ + front vowel in initial and intervocalic position[90]
BranchLanguageki, keɡi, ɡe
initialmedialinitialmedial
SardinianCampidaneset͡ʃʒd͡ʒ∅
Central Sardiniankɡɣ
Logudoresekɣɡ∅
Western RomanceWest/North Friuliant͡ʃʒ(d)ʒ/zj/ʒ
East Friulianszʒ/zj/ʒ
Fassant͡ʃʒ
Comelicanθðd/ð
Livinallongheset͡ʃʒ
Surselvant͡ʃʒɟ
Sutselvant͡ʃʒɟ(ʒ)
Surmirant͡ʃʒd͡ʒd͡z
Engadinian(t)ʃʒɟ/j
Venetianθz(d)z/ðz/ð
Ligurians(t͡s)ʒz(d͡z)
Lombard(t)ʃ/sz(d)ʒ/zʒ
Picardʃz(ʒ)ɡj
Frenchsjzʒj
Franco-Provençals(θ)z(ʒ)d͡z(z)j
Auvergnatszd͡z(d)ʒ/d͡z
Occitanszʒ
Catalansz/ð[lower-roman 2](d)ʒ
Spanishθ[lower-roman 3]j/∅
Portugueseszʒʒ/∅
OtherRomaniant͡ʃd͡ʒ
Vegliotet͡ʃi, ked͡ʒi, ɡed͡ʒi, ɡ(e)
Calabreset͡ʃj
Tuscanʃd͡ʒdːʒ
Corsicant͡ʃɟ/d͡ʒɟ/j

/k ɡ/ + /a/

In some Gallo-Romance languages, /k ɡ/ came to be palatalized before original /a/.[79] This is assumed to have taken place more recently than palatalization before high and mid front vowels,[79] and can have a different outcome. Palatalization and affrication of /k/ before /a/ occurs in all central French dialects,[91] but not in northern varieties (Norman and Picard)[91] beyond the Joret line. Nevertheless, forms such as the Picard kièvre, kier < CAPRAM 'goat', CARUM 'dear' do show a degree of palatalization before fronted outcomes of Latin /a/ (coarticulation but not affrication).[92] Accordingly, it has been proposed that this was the original environment for palatalization in other French dialects as well, at a time when the fronting of original /a/ in this environment was still allophonic, and that it later spread by analogy to any velar + /a/ sequence. An alternative theory holds that /a/ may have been a front vowel in general, as it happens to be in modern French, where as a result the initial consonant of e.g. quatre 'four' may be palatalized to [kʲ] or [c].[93]

In French, original /k/ before /a/ developed to ⟨ch⟩ (pronounced [t͡ʃ] in Old French and [ʃ] today),[94] as in CANTARE 'sing' > chanter[95] /ʃɑ̃te/. Thus, the palatalized outcome of /k/ in /ka/ remains distinct in French from the outcome of /k/ before /e i j/, as in CENTUM 'hundred' > cent /sɑ̃/. There was a parallel change of /ɡ/ before /a/ to ⟨j⟩ (pronounced [d͡ʒ] in Old French and [ʒ] today),[94] as in GAMBAM > jambe[96] /ʒɑ̃b/. This palatalization must have preceded the monophthongization of Latin AU to French o, based on outcomes such as CAUSAM > chose and GAUDIA > joie.[94]

Palatalization of /ka ɡa/ is also found in Occitan, where it is attested from the earliest texts in the language. In general, northern dialects have it and southern ones do not, but neither group is entirely uniform in this regard as the geographic extent of palatalization is subject to lexical variation.[97] The overall distribution of palatalization, with a clear weakening from north to south, as well as the presence of toponyms in the north that show retention of /ka ɡa/, suggest that the phenomenon was historically imported into Occitan from French dialects.[98] The Occitan outcomes of palatalized /k ɡ/ before /a/ are variously /t͡ʃ d͡ʒ/, /t͡s d͡z/, /s z/, or rarely /θ ð/.[99] Examples include the Lemosin [d͡ʒaˈlinɔ] < GALLINAM 'hen' and southern Auvergnat [t͡sasˈtɛ(r)] < CASTELLUM 'castle'.[100]

Aside from Gallo-Romance, palatalization of /ka ɡa/ is found broadly in Rhaeto-Romance[lower-roman 5] and also, in widely scattered traces, in the dialects of northern Italy (Gallo-Italic and Venetian).[79][102] This is sometimes interpreted as having a common origin with the northern Gallo-Romance palatalization, though it may instead be a parallel but independent development.[103] Some varieties of Friulian show the affricates [t͡ʃ d͡ʒ] (as in CABALLUM > [t͡ʃaˈval] 'horse', GAMBAM > [ˈd͡ʒambe] 'leg'), while in central and northern Friulian the palatal stops [c ɟ] are found instead.[79][104]

Velar + coronal

In addition to palatalization of consonants followed by /j/ or by front vowels, some Romance languages show palatalized outcomes for Latin consonant clusters, particularly ones where a velar consonant was followed by a coronal one.

/ɡn/

The most widespread outcome of the Latin sequence -GN- is the palatal nasal [ɲ(ː)],[105] which merged with the outcome of original /nj/. This development is found throughout Western Romance[15] (e.g. Spanish [ˈpuɲo], Portuguese [ˈpuɲu] Catalan [ˈpuɲ] from Latin PUGNUM[106]) and in Tuscan Italian. (This merger seemingly postdates the raising of the vowel derived from Latin short -Ĭ- in a stressed syllable before original /nj/, as in GRAMĬNEAM > Italian gramigna, since unraised [e] is found in Italian legno < LĬGNUM.[107])

A minority[108] of languages instead show vocalization of the velar consonant:

  • [-jn-] is attested as an outcome in some languages of central and southern Italy.[106]
  • [-wn-] is attested as an outcome in some Southern Italian languages, as in [ˈliwna] 'firewood' < LIGNA.[106] This is found in a more limited area (comprising Basilicata, central southern Puglia, and central northern Calabria).[108]

Non-palatalized outcomes of -GN- are found in Romanian, where it developed to [-mn-] (as in LIGNUM > [ˈlemn] 'wood'),[106] and in Sardinian, where it developed to [-nn-] (as in LIGNUM > [ˈlinnu]).[106]

Some loans into Albanian show an outcome of nj, such as denjë < DIGNUM and shenjë < SIGNUM, while others show the same development as plain n (vgje ~ vgjê < ABIEGNUM and kunat ~ kunatën < COGNATUM) and a third outcome is seen in peng < PIGNUM.[109]

/kt/

Western Romance languages show a change of Latin -CT- /kt/ to [jt], which could develop further into an affricate such as [t͡ʃ].[15]

Some Romance languages have non-palatalized outcomes of this cluster, such as Romanian (where -CT- developed to [pt], as in [ˈfapt] < FACTUM) and Italian (where -CT- developed to [tt], as in [ˈfatto] < FACTUM).[106]

In Albanian, some words loaned from Romance/Latin show -CT- > jt (such as DIRECTUM > drejtë), which Orel attributes to borrowing from a West Balkan variety showing the same development as in Western Romance, whereas others show -CT- > ft (such as LUCTAM > luftë), with the velar changed to a labial as in Romanian.[110]

/ks/

Western Romance languages show development of Latin -X- /ks/ to [js], which fell together with the outcome of /ssj/ and shows various final outcomes including [ʃ].[15]

Some Romance languages have non-palatalized outcomes of this cluster.

  • The usual outcome of /ks/ in Italian is /ss/, as in [ˈasse] < AXEM.[106] In some isolated cases,[105] Italian appears to show development of /ks/ to [ʃʃ]. It is unclear whether both outcomes are indigenous Tuscan developments.[111] Some cases of [ʃʃ] can be explained as actually developing from metathesis of /ks/ to /sk/, which was then regularly palatalized by a following front vowel, as in the case of ascella, which can be traced to ASCELLA, ASCILLA, attested Late Latin variants of AXILLA.[105] Others, such as coscia from Latin COXAM, may have been borrowed from a non-Tuscan variety, such as Old French or Provençal.[105] However, the words in which [ʃ] occurs in Tuscan do not appear to otherwise show signs of being words of exotic or north-western origin.[111]
  • Romanian generally shows /ps/, as in [ˈkwapsə] < COXAM,[106] but also has /s/ as a common alternative outcome, as in [ləˈsa] < LAXARE.[112]

/kl ɡl/

In Ibero-Romance and Gallo-Romance, intervocalic /kl ɡl/ (including cases where these developed via syncope from the Classical Latin sequences -CUL-, -TUL-, -GUL-) evolved into /ʎ/, thus merging with the outcome of original /lj/.[113] However, a number of dialects of Ladin do not show this palatalization.[114]

In Italian and Romanian, these developed differently.[113] Intervocalic /kl/ in these languages has outcomes that show loss of lateral articulation rather than loss of the original stop, as in Latin OCULUM > Italian occhio 'eye' (with /kkj/[113]) and Latin AURICULAM > Romanian ureche 'ear' (with /c/[113]).

There are competing explanations of the change from /kl ɡl/ to [ʎ]; one hypothesis supposes the series of changes /kl/ > [kʎ] > [çʎ] > [ʎ], while according to another hypothesis the development was /kl/ > [çl] > [jl] > [jʎ] > [ʎ].[115]

Obstruent + /l/

The Latin sequences /pl bl fl kl ɡl/ yield palatalized reflexes in numerous Romance languages. This probably began with /l/ allophonically turning to [ʎ] after a velar consonant; the resulting system *[pl bl fl kʎ ɡʎ] underlies Balkan Romance, northern Abruzzese, old Gallo-Italic, and old Venetian.[116]

Controversially, the outcomes in most of Gallo-Romance, and Catalan, can also be traced to the same underlying system if one assumes that there followed, for phonological reasons, a reversion of /kl ɡl/ *[kʎ ɡʎ] to [kl ɡl] in fortis positions[lower-roman 6] after the lenition of *[kʎ ɡʎ] to [ʎ] in lenis positions.[lower-roman 7] This is at odds with the traditional view that Latin [kl ɡl] remained unchanged in fortis positions all along.[117]

The outcomes in Italo-Romance (other than northern Abruzzese) can be traced to a system *[pʎ bʎ fʎ kʎ ɡʎ] that probably developed from the older system mentioned above via generalization of post-obstruent [ʎ].[118] The same is true for most Franco-Provençal dialects.[119] Ibero-Romance seems consistent with this as far as *[pʎ fʎ kʎ] and word-internal *[ɡʎ bʎ] are concerned, but it shows non-palatalized outcomes of Latin word-initial [bl ɡl]. It has been proposed that the inherited words that began with the latter sequences were early on 'submerged' by learned borrowings from Latin.[120]

Morphological consequences

Palatalization gave rise to various morphological effects across the Romance languages.

In verb conjugation, the original presence of either [j] or of a front vowel in some forms but not in others resulted in patterns of alternation between different stems for different person-number combinations. These alternations were frequently subject to morphological leveling, but could also be extended by analogy to verbs with different etymologies; both of these competing tendencies resulted in the creation of verb forms that are different from those predicted by applying regular sound changes to the corresponding Latin forms.

The outcomes of the verb COLLIGO, COLLIGERE, discussed above, provide examples of leveling and analogical extension.

In Spanish, Latin COLLIGO initially developed to cuelgo,[87] but this was later changed under the influence of coge to coxgo,[87] which in modern Spanish has been fully leveled to cojo.

In Italian, the [lg] found in the forms colgo and scelgo, from COLLIGO and *EXELIGO, was extended by analogy[121] to some verb forms that originally had [lj], such as DOLEO > doglio (by regular sound change) and dolgo (analogical), SALIO > salgo (by analogy),[122] also valgo.[121]

Spelling of palatalized consonants
For the spellings of palatalized n and l in the Romance languages, see palatalization (sound change) § Mouillé.

The effects of palatalization are seen in the writing systems of almost all Romance languages, where the letters ⟨c, g⟩ have the "hard" pronunciations [k ɡ] in most situations, but "soft" pronunciations (e.g. French/Portuguese [s, ʒ], Italian/Romanian [t͡ʃ, d͡ʒ]) before ⟨e, i, y⟩.[123] (This orthographic trait has passed into Modern English through Norman French-speaking scribes writing Middle English, replacing the earlier system of Old English.) This has the effect of keeping the modern spelling similar to the original Latin one but complicates the relationship between sound and letter. In particular, the hard sounds must be written differently before ⟨e, i, y⟩ (e.g. Italian ⟨ch, gh⟩, Portuguese ⟨qu, gu⟩), and likewise for the soft sounds when not before these letters (e.g. Italian ⟨ci, gi⟩, Portuguese ⟨ç, j⟩). Furthermore, in Spanish, Catalan, Occitan and Brazilian Portuguese, the use of digraphs containing ⟨u⟩ to signal the hard pronunciation before ⟨e, i, y⟩ means that a different spelling is also needed to signal the sounds /kw ɡw/ before these vowels (Spanish ⟨cu, gü⟩, Catalan, Occitan and Brazilian Portuguese ⟨qü, gü⟩).[lower-roman 8] This produces a number of orthographic alternations in verbs whose pronunciation is entirely regular. The following are examples of corresponding first-person plural indicative and subjunctive in a number of regular Portuguese verbs: marcamos, marquemos "we mark"; caçamos, cacemos "we hunt"; chegamos, cheguemos "we arrive"; averiguamos, averigüemos "we verify"; adequamos, adeqüemos "we adapt"; oferecemos, ofereçamos "we offer"; dirigimos, dirijamos "we drive" erguemos, ergamos "we raise"; delinquimos, delincamos "we commit a crime".

In the case of Italian, the convention of digraphs ⟨ch⟩ and ⟨gh⟩ to represent /k ɡ/ before written ⟨e, i⟩ results in similar orthographic alternations, such as dimentico 'I forget', dimentichi 'you forget', baco 'worm', bachi 'worms' with [k] or pago 'I pay', paghi 'you pay' and lago 'lake', laghi 'lakes' with [ɡ]. The use in Italian of ⟨ci⟩ and ⟨gi⟩ to represent /t͡ʃ/ or /d͡ʒ/ before vowels written ⟨a⟩, ⟨o⟩, ⟨u⟩ neatly distinguishes dico 'I say' with /k/ from dici 'you say' with /t͡ʃ/ or ghiro 'dormouse' /ɡ/ and giro 'turn, revolution' /d͡ʒ/, but since orthographic ⟨ci⟩ and ⟨gi⟩ are also used to represent sequences of /t͡ʃ/ or /d͡ʒ/ and the vowel phoneme /i/ (as in dici /ˈdit͡ʃi/, giro /ˈd͡ʒiro/), the status of ⟨i⟩ followed by another vowel in spelling can be ambiguous. For example, the written forms offer no indication that ⟨cia⟩ in camicia 'shirt' represents a single unstressed syllable /t͡ʃa/, (/kaˈmit͡ʃa/ [kaˈmiːt͡ʃa~kaˈmiːʃa]), but that ⟨cia⟩ in farmacia 'pharmacy' is a bisyllabic sequence consisting of the stressed syllable /t͡ʃi/ and syllabic /a/ (/farmaˈt͡ʃia/ [farmaˈt͡ʃiːa~farmaˈʃiːa]).

Notes
  1. In this article Latin lexemes will be cited in their normal spelling, even in cases where the cited source writes them in some modified way, for instance without final ⟨M⟩. They are written in small capitals, as is often the practice in Romance linguistics, since this provides a convenient way to indicate Latin words without having to repeat the word 'Latin'.
  2. Subsequently changed further: in modern Catalan, Latin intervocalic -TJ- and Latin -C- followed by E or I were ultimately lost between vowels (e.g. RATIONEM > raó; VECINUM > veí) and vocalized to /w/ word-finally or before a consonant (e.g. PALATIUM > palau, PUTEUM > pou, PRETIUM > preu; DECEM > deu, VICEM > veu). This is identical to the outcome of Latin intervocalic -D-.
  3. In most of European Spanish; Latin America instead has /s/.
  4. In Spanish phonology, the palatal semivowel [j] is usually analyzed in this context as a non-syllabic allophone of the vowel /i/.
  5. Everywhere except Western Romansh, where the palatalization only occurs before stressed /a/. This seems to reflect the original environment from which Eastern Romansh, and perhaps the rest of Rhaeto-Romance, later extended the phenomenon to unstressed syllables. Curiously, in the Western Romansh (Sutselvan) dialect of Ems, the original velar /k/ appears to have been restored before stressed /a/.[101]
  6. Post-pausal or post-consonantal. Most commonly word-initial, even more so after the Gallo-Roman loss of most final vowels reduced the incidence of word-initial intervocalic contexts.
  7. Intervocalic, commonly word-internal.
  8. The current Portuguese spelling has abolished the use of the diaeresis for this purpose.

References
  1. Repetti 2016, p. 658.
  2. Barbato 2022, §1.
  3. Repetti 2016, p. 659.
  4. Recasens 2020, §3.1.1.
  5. Kerkhof 2018, p. 128.
  6. Kerkhof 2018, p. 129.
  7. Orel 2000, p. 108.
  8. Carnoy 1916, p. 146.
  9. Barbato 2022, §6.
  10. Barbato 2022, §2.
  11. Barbato 2022, §3.
  12. Recasens 2020, §3.5.
  13. Kerkhof 2018, p. 127.
  14. Boyd-Bowman 1980, p. 130.
  15. Barbato 2022, §5.
  16. Kerkhof 2018, p. 127-130.
  17. Recasens 2020, §3.1.1, Table 1a.
  18. Maiden 2013, p. 53.
  19. Maiden 2013, p. 87.
  20. Maiden 2013, p. 54.
  21. Maiden 2013, p. 54-55.
  22. Kutzner 1982, pp. 84–87.
  23. Kutzner 1982, pp. 85.
  24. Mackenzie 1999–2022, "History of Spanish Consonants", §2.2, 3.2, 4.1.
  25. Penny 2002, pp. 64–65.
  26. Penny 2002, p. 65.
  27. Boyd-Bowman 1980, p. 71.
  28. Recasens 2020, §3.1.1, Table 1a; §3.1.2, Table 2; §3.3, Table 4b.
  29. Barbato 2022, §4.
  30. Barbato 2022, §1, 5.
  31. Zampaulo 2019, p. 78.
  32. Nyrop 1914, p. 422.
  33. Zampaulo 2019, p. 79.
  34. Bourciez 1921, p. 253.
  35. Repetti 2016, p. 661.
  36. Burger 1955, p. 24.
  37. Zampaulo 2019, p. 78-79.
  38. Brittain 1900, p. 64.
  39. Pope 1934, p. 229.
  40. Boyd-Bowman 1980, p. 89.
  41. Repetti 2016, p. 662.
  42. Barbato 2022, §3, 4, 5.
  43. Zampaulo 2019, pp. 50, 58–62.
  44. Kutzner 1982, pp. 84–85.
  45. Repetti 2016, p. 660.
  46. Operstein 2010, p. 118.
  47. Wireback 2002, p. 311.
  48. Operstein 2010, p. 114.
  49. Canalis 2017, p. 157.
  50. Canalis 2017, p. 162.
  51. Canalis 2017, p. 161.
  52. Canalis 2017, p. 163.
  53. Barbato 2022, §2, 3.
  54. Recasens 2020, §2.1.4.
  55. Repetti 2016, p. 661-662.
  56. Boyd-Bowman 1980, p. 107.
  57. Brittain 1900, p. 63.
  58. Lloyd 1987, p. 264.
  59. Penny 2002, p. 84, 272.
  60. Wireback 2002, p. 312.
  61. Wireback 2002, p. 321.
  62. Wireback 2002, p. 311-312.
  63. Wireback 2002, p. 322-326.
  64. Mackenzie 1999–2022, History of the Spanish Verb", §6.
  65. Operstein 2010, p. 109.
  66. Operstein 2010, p. 110.
  67. Boyd-Bowman 1980, p. 106.
  68. Brittain 1900, p. 62.
  69. Operstein 2010, p. 115.
  70. Kerkhof 2018, p. 149.
  71. Repetti 2016, p. 662-663.
  72. Repetti 2016, p. 663.
  73. Merlo 2014.
  74. Merlo 2014, p. 171.
  75. Maiden 2016, p. 127.
  76. Loporcaro 2009, p. 145.
  77. Kerkhof 2018, p. 139.
  78. Maiden 2013, p. 58.
  79. Repetti 2016, p. 664.
  80. Penny 2002, p. 69.
  81. Mackenzie 1999–2022, "History of Spanish Consonants", §3.2.
  82. Maiden 2013, p. 59.
  83. Rohlfs 1966, p. 377.
  84. Boyd-Bowman 1980, p. 77.
  85. Hanssen 1913, p. 99.
  86. Penny 2002, p. 49.
  87. Tuttle 1912, p. 422.
  88. Malkiel 1982, p. 249.
  89. Recasens 2020.
  90. Recasens 2020, §3.3, Table 4a.
  91. Kerkhof 2018, p. 150.
  92. Buckley 2009, p. 56–57.
  93. Buckley 2009.
  94. Boyd-Bowman 1980, p. 30.
  95. Boyd-Bowman 1980, p. 29.
  96. Boyd-Bowman 1980, p. 31.
  97. Buckley 2009, p. 57–58.
  98. Buckley 2009, p. 58.
  99. Buckley 2009, p. 58, §6.
  100. Mooney 2022, p. 716.
  101. Haiman & Benincà 1992, §1.2.3.1a.
  102. De Cia & Iubini-Hampton 2020, §3.
  103. Kerkhof 2018, pp. 150–151.
  104. De Cia & Iubini-Hampton 2020, §2.1.
  105. Baglioni 2014, p. 10.
  106. Repetti 2016, p. 667.
  107. Baglioni 2014, p. 14.
  108. Baglioni 2014, p. 15.
  109. Orel 2000, p. 115.
  110. Orel 2000, p. 116.
  111. Maiden 2013, p. 57.
  112. Sala 1976, pp. 171–185.
  113. Zampaulo 2019, p. 51.
  114. Repetti & Tuttle 1987, p. 61.
  115. Recasens 2020, §4.3.
  116. Repetti & Tuttle 1987, §§1.0–1.1.
  117. Repetti & Tuttle 1987, §1.2.
  118. Repetti & Tuttle 1987, §2.0.
  119. Repetti & Tuttle 1987, §2.2.
  120. Repetti & Tuttle 1987, §§2.5–2.6.
  121. Grandgent 1927, p. 122.
  122. Grandgent 1927, p. 106.
  123. Boyd-Bowman 1980, p. 33.

Bibliography
  • Baglioni, Daniele (2014). Molinelli, Piera; Cuzzolin, Pierluigi; Fedriani, Chiara (eds.). "Il nesso GN dal latino alle lingue romanze: questioni aperte e prospettive di ricerca". Latin vulgaire, latin tardif X: Actes du Xe colloque international sur le latin vulgaire et tardif. Bergamo University Press: Sestante Edizioni. 1: 3–24.
  • Barbato, Marcello (20 June 2022). "The Early History of Romance Palatalizations". oxfordre.com. Retrieved 11 September 2023.
  • Bourciez, Édouard (1921). Précis historique de phonétique française (5th ed.). Paris: Librairie C. Klincksieck.
  • Boyd-Bowman, Peter (1980). From Latin to Romance in Sound Charts. Washington, D.C.: Georgetown University Press. ISBN 9780878400775.
  • Brittain, Margaret S. (1900). Historical Primer of French Phonetics and Inflection. Oxford: Clarendon Press.
  • Buckley, Eugene (2009). "Phonetics and phonology in Gallo-Romance palatalisation". Transactions of the Philological Society. 107: 31–65.
  • Burger, André (1955). "Phonématique et Diachronie a Propos de La Palatalisation Des Consonnes Romanes". Cahiers Ferdinand de Saussure. 13: 19–33. Retrieved 12 September 2023.
  • Canalis, Stefano (2017). "The voiced and voiceless outcomes of intervocalic -sj- in Old Tuscan". Quaderni di Linguistica e Studi Orientali / Working Papers in Linguistics and Oriental Studies. 3: 157–182.
  • Carnoy, Albert J. (1916). "Some Obscurities in the Assibilation of ti and di before a Vowel in Vulgar Latin". Transactions and Proceedings of the American Philological Association. 47: 145–152. Retrieved 16 September 2023.
  • De Cia, Simone; Iubini-Hampton, Jessica (2020). "Debunking Rhaeto-Romance: Synchronic evidence from two peripheral Northern Italian dialects". Modern Languages Open. 7 (1): 1–18.
  • Grandgent, Charles H. (1927). From Latin to Italian: An Historical Outline of the Phonology and Morphology of the Italian Language. Cambridge: Harvard University Press.
  • Haiman, John; Benincà, Paola (1992). The Rhaeto-Romance languages. London: Routledge.
  • Hanssen, Federico (1913). Gramática histórica de la lengua castellana (in Spanish). Halle, Max Niemeyer.
  • Kerkhof, P.A. (9 October 2018). Language, law and loanwords in early medieval Gaul: language contact and studies in Gallo-Romance phonology (Thesis). Retrieved 11 September 2023.
  • Kutzner, Evelyn Beatrice (May 1982). A synchronic study of dental consonants preceding yod in French and its implications for a diachronic analysis (Thesis).
  • Loporcaro, Michele (2009). "Chapter 3: Phonological processes". In Martin, Maiden; Smith, John Charles; Ledgeway, Adam (eds.). The Cambridge History of the Romance Languages. Cambridge University Press. pp. 109–154.
  • Lloyd, Paul M. (1987). From Latin to Spanish: Historical phonology and morphology of the Spanish Language. Memoirs of the American Philosophical Society. American Philosophical Society.
  • Mackenzie, Ian (1999–2022). "The Linguistics of Spanish". Retrieved 12 September 2023.
  • Maiden, Martin (2013) [1st pub. 1995]. Robins, R. H.; Harris, Martin; Horrocks, Geoffrey (eds.). A Linguistic History of Italian. Longman Linguistics Library. Routledge.
  • Maiden, Martin (2016). "Chapter 9: Dalmatian". In Ledgeway, Adam; Maiden, Martin (eds.). The Oxford Guide to the Romance Languages. Oxford University Press. pp. 658–668.
  • Malkiel, Yakov (1982). "Interplay of Sounds and Forms in the Shaping of Three Old Spanish Medial Consonant Clusters". Hispanic Review. 50 (3): 247–266.
  • Merlo, Roberto (2014). "Un problema trascurato di fonetica storica romena: la "palatalizzazione delle velari" tra eredità latina e interferenze slave" (PDF). Dacoromania. 19: 165–197. Retrieved 15 September 2023.
  • Mooney, Damien (2022). "Chapter 22: Occitan". In Gabriel, Christoph; Gess, Randall; Meisenburg, Trudel (eds.). Manual of Romance Phonetics and Phonology. Berlin, Boston: De Gruyter. pp. 709–742.
  • Nyrop, Kristoffer (1914). Grammaire historique de la langue française. Vol. 1 (3rd ed.). Gyldendalske Boghandel, Nordisk Forlag.
  • Operstein, Natalie (2010). Consonant Structure and Prevocalization. Amsterdam Studies in the Theory and History of Linguistic Science. Series IV, Current Issues in Linguistic Theory 312. John Benjamins Publishing Company.
  • Orel, Vladimir (2000). A Concise Historical Grammar of the Albanian Language: Reconstruction of Proto-Albanian. Brill.
  • Penny, Ralph (2002). A History of the Spanish Language (2nd ed.). Cambridge, United Kingdom: Cambridge University Press.
  • Pope, Mildred K. (1934). From Latin to modern French. Manchester, United Kingdom: Manchester University Press.
  • Recasens, Daniel (30 July 2020). "Palatalizations in the Romance Languages". oxfordre.com. Retrieved 11 September 2023.
  • Repetti, Lori; Tuttle, Edward F. (1987). "The Evolution of Latin PL, BL, and CL, GL in Western Romance". Studi mediolatini e volgari. 33: 53–115.
  • Repetti, Lori (2016). "Chapter 39: Palatalization". In Ledgeway, Adam; Maiden, Martin (eds.). The Oxford Guide to the Romance Languages. Oxford University Press. pp. 658–668.
  • Rohlfs, Gerhard (1966). Grammatica storica della lingua italiana e dei suoi dialetti: fonetica. Turin: Giulio Einaudi.
  • Sala, Marius (1976). Contributions à la phonétique historique du roumain. Paris: Klincksieck.
  • Tuttle, E. H. (1912). "Colligere in Spanish". Romanic Review, The. Columbia University Press. 3: 422–423.
  • Wireback, Kenneth J. (2002). "On the Metathesis of Labials + /j/ in Hispano-Romance". Hispanic Review. 70 (3): 311–331.
  • Zampaulo, André (2019). Palatal Sound Change in the Romance Languages: Diachronic and Synchronic Perspectives. Oxford Studies in Diachronic and Historical Linguistics. Oxford University Press.
This article is issued from Wikipedia. The text is licensed under Creative Commons - Attribution - Sharealike. Additional terms may apply for the media files.