言語史を述べる場合、記述を「外的な歴史」と「内的な歴史」に区分することが通例であり、本稿もそれに従う。外的な歴史 external history は民族・社会・政治・技術などの変化が言語に及ぼす影響を論じるものであり、内的な歴史 internal history は自発的要因から言語が被る音韻や文法の変化を論じるものである。
The syntax shows the systematic presence of a subject pronoun in front of the verb, as in the Germanic languages: je vois, tu vois, il voit, while the subject pronoun is optional – function of the parameter pro-drop – in the other Romance languages (as in veo, ves, ve).
疑問文における主語と述語の倒置。これはゲルマン諸語の特徴であり、フランス語以外の主要なロマンス諸語には見られない(Vous avez un crayon.→Avez-vous un crayon?(鉛筆を持っていますか。))。
名詞の前に形容詞を置くのは、ゲルマン諸語の特徴で、他の主要なロマンス諸語よりもフランス語で一般的で、ときにはそれが必須となっている(belle femme, vieil homme, grande table, petite table)。任意の場合は、その意味が変わってくる(grand homme(「偉人」)とle plus grand homme(「最も偉大な人」)/homme grand(「背の高い人」)とl'homme le plus grand(「最も背の高い人」)、certaine chose/chose certaine)。ワロン語では《形容詞+名詞》の語順は、古フランス語のように一般的な規則である。
中世イタリアの詩人ダンテはその著書「俗語論」でロマンス諸語は「はい」(現在の標準フランス語ではoui)と言うのに用いる単語によって3つの類例に分類できるとした。
Nam alii oc, alii si, alii vero dicunt oil(ocと言う人もいれば、siと言う人もいれば、oïlと言う人もいる)
ラテン語のhocille「それはそれ」に由来するoïlは北フランスを、ラテン語のhoc「それ」に由来するocは南フランスを、ラテン語のsic「従って」に由来するsiはイタリア半島やイベリア半島を占めた。現代の言語学者は、概して現代語でouèとなるリヨン周辺のフランスにおける第3類「アルピタン語」を加えている。
行政機関として初めてフランス語を公用語として採用したのはイタリア北西部のヴァッレ・ダオスタで、1536年のことであったが、これはフランスによるフランス語公用語化に3年先立つものである[9]。1539年のヴィレル=コトレ勅令でフランソワ1世はフランス語を行政と宮廷で用いる公用語とし、それ以前に用いられていたラテン語を追放した。公的機関で用いるべき標準語として使用を強制されたことと、曲用体系を失ったことをもって、オイル語のこの方言は古フランス語と区別される中世フランス語(moyen français)とされている。1550年にはフランス語文法について最初に記述したルイ・メグレ(フランス語版)のTretté de la Grammaire françaiseが出版されている。現代フランス語で700語を数える、美術(scenario、piano)・嗜好品・食品などを表すイタリア語起源の語彙がこの時期に持ち込まれた[10]。
A protected /j/ (not preceded by a vowel), stemming from an initial /j/ or from a /dj/, /ɡj/, or /ɡ(eˌi)/ when preceded by a consonant, becomes /dʒ/.
A /j/ followed by another consonant tends to palatalize that consonant; these consonants may have been brought together by intertonic loss. (E.g. medietate > /mejetate/ > /mejtʲate/ > 'moitié'. peior > /pejro/ > /piejrʲe/ > 'pire', but impeiorare > /empejrare/ > /empejrʲare/ > /empejriɛr/ > OF 'empoirier' "to worsen".)
Palatalized sounds lose their palatal quality and eject a /j/ into the end of the preceding syllable, when open; also into the beginning of the following syllable when it is stressed, open, and front (i.e. /a/ or /e/). Hence *cugitare > /kujetare/ > /kujdare/ > /kujdʲare/ >> /kujdiɛr/ OF 'cuidier' "to think". mansionata > /mazʲonada/ > /mazʲnada/ > /majzʲnjɛðə/ > OF 'maisniée' "household".
/tʃ/ and /dʒ/ (including those from later sources, see below) eject a following /j/ normally, but do not eject any preceding /j/.
Double /ssʲ/ < /ssj/ and from various other combinations also ejects a preceding /j/.
Single /dz/ ejects such a /j/, but not double /tts/, evidently since it is a double sound and causes the previous syllable to close; see comment above, under lenition.
Actual palatal /lʲ/ and /nʲ/ (as opposed to the merely patalized varieties of the other sounds) retain their palatal nature and don't emit preceding /j/. Or rather, palatal /lʲ/ does not eject a preceding /j/ (or else, it is always absorbed, even when depalatalized); palatal /nʲ/ emits a preceding /j/ when depalatalized, even if the preceding syllable is closed, e.g. jungit > *yōnyet > /dʒoɲt/ > /dʒojnt/ 'joint'.
Palatal /rʲ/ ejects a preceding /j/ as normal, but the /j/ metathesizes when a /a/ precedes, hence operariu > /obrarʲo/ > /obrjaro/ (not [*/obrajro/]) >> 'ouvrier' "worker".
Second diphthongization: diphthongization of /e/, /o/, /a/ to /ei/, /ou/, /ae/ in stressed, open syllables, not followed by a palatal sound (not in all Gallo-Romance). (Later on, /ei/ > /oi/, /ou/ > /eu/, /ae/ > /e/; see below.)
Second unstressed vowel loss: Loss of all vowels except /a/ in unstressed, final syllables; addition of a final, supporting /e/ when necessary, to avoid words with impermissible final clusters.
Second lenition: Same changes as in first lenition, applied again (not in all Gallo-Romance). NOTE: Losses of unstressed vowels may have blocked this change from happening.
Palatalization of /ka/ > /tʃa/, /ɡa/ > /dʒa/.
Further vocalic changes (part 1):
/ae/ > /ɛ/ (but > /jɛ/ after a palatal, and > /aj/ before nasals when not after a palatal).
/au/ > /ɔ/.
Further consonant changes:
Geminate stops become single stops.
Final stops and fricatives become devoiced.
/dz/ > /z/, when not final.
A /t/ is inserted between palatal /ɲ/, /ʎ/ and following /s/ (doles > 'duels' "you hurt" but colligis > *colyes > 'cuelz, cueuz' "you gather"; jungis > *yōnyes > 'joinz' "you join"; filius > 'filz' "son").
Palatal /ɲ/, /ʎ/ are depalatalized to /n/, /l/ when final or following a consonant.
In first-person verb forms, they may remain palatal when final due to the influence of the palatalized subjunctives.
/ɲ/ > /jn/ when depalatalizing, but /ʎ/ > /l/, without a yod. (*veclus > /vɛlʲo/ > /viɛlʲo/ > 'viel' "old" but cuneum > /konʲo/ > 'coin'. balneum > /banjo/ > 'bain' but montanea > /montanja/ > 'montagne'.)
Diphthongs are consistently rendered as falling diphthongs, i.e. the major stress is on the first element, including for /ie/, /ue/, /ui/, etc. in contrast with the normal Spanish pronunciation.
NOTE: Changes here affect oral and nasal vowels alike, unless otherwise indicated.
/o/ > /u/.
/l/ before consonant becomes /w/.
/ue/ and /eu/ > /œ/.
Rising diphthongs develop when first element of diphthong is /u/, /y/ or /i/, causing the stress to shift to the second element in these cases (hence /yi/[yj] > [ɥi]).
/oi/ > /we/. This in turn develops to /ɛ/ in some words, e.g. français; note doublet François. Much later, perhaps in the 17th century, remaining /we/ sounds > /wa/ except in "court" pronunciation. (The /wa/ pronunciation was then stigmatized as "vulgar" until the French Revolution but remaining more or less in use in Quebec.) However, nasalized /wẽ/ was unaffected; hence ModF 'coin' "corner" /kwɛ̃/not **/kwɑ̃/.
/ai/ merges into /ɛ/; after this, 'ai' is a common spelling of /ɛ/, regardless of origin. ('è' is a later development.)
/e/ merges into /ɛ/ in closed syllables.
/ts/ > /s/, /tʃ/ > /ʃ/, /dʒ/ > /ʒ/.
Loss of /s/ before any consonant, with lengthening of preceding vowel. This may have begun as early as 900 AD or so, when /s/ before a consonant became /h/. Later on the /h/ vanished with compensatory lengthening of the preceding vowel. From borrowings into English, it appeared that this latter stage had already occurred in Old French when the following consonant was voiced but not when it was unvoiced. By the end of Old French, the latter stage was complete and a whole new set of phonemically lengthened vowels developed. These were still marked in writing with an 's', but starting around 1700 were marked instead with circumflex over the vowel (perhaps because actual pronounced /s/ had been reintroduced into that position in certain words, e.g. due to borrowing of learned words from Latin.)
Development of two low vowels /a/ and /ɑ/. The latter was initially an allophone of /a/ that occurred before /s/ and /z/, and become phonemic when /ts/ merged with /s/. (e.g. Mod. Fr. 'chasse' /ʃas/ "(he) hunts" < [*/cattsa/] < captiat vs. 'châsse' /ʃɑs/ "reliquary, (eyeglass) frame" < [*/cassa/] < capsa "strong box".) Later losses of /s/ produced further minimal pairs, e.g. 'pâte' /pɑt/ "paste" < VL *pasta vs. 'patte' /pat/ "paw" < VL *patta; or 'bas' /bɑ/ "low" < /bas/ < bassum vs. 'bat' /ba/ "(he) beats" < /bat/ < VL *battet < battuet.)
中世フランス語(1500年代)まで
NOTE: Changes here affect oral and nasal vowels alike, unless otherwise indicated.
/au/ > /o/.
/ei/ > /ɛ/.
Loss of final consonants before a word beginning with a consonant. This produces a three-way pronunciation for many words (alone, followed by a vowel, followed by a consonant), which is maintained to this day in the words 'six' "six" and 'dix' "ten" (and until recently 'neuf' "nine"), e.g. 'dix' /dis/ "ten" but 'dix amis' /diz ami/ "ten friends" and 'dix femmes' /di fam/ "ten women".
(Around this time, subject pronouns become mandatory.)
'oi' /we/ > /wa/ (上記を参照 – Through late Old French) or /ɛ/ (e.g. étoit > était – 19th c.).
(fill in further)
現代フランス語(2000年代)まで
/r/ becomes uvular sound: trill /ʀ/ or fricative /ʁ/, (replacing the rolled 'r' formerly often used by the clergy).
Loss of final /ə/. Loss of /ə/ elsewhere unless a sequence of three consonants would be produced (such constraints operate over multiword sequences of words that are syntactically connected).
Gradual loss of liaison
Gradual loss of the "ne" in negations, "je n'ai pas" becomes "j'ai pas".
(fill in further)
鼻音化
Progressive nasalization of vowels before /n/ or /m/ occurred over several hundred years, beginning with the low vowels, possibly as early as c. 900 AD, and finished with the high vowels, possibly as late as c. 1300 AD. Numerous changes occurred afterwards, continuing up through the present day.
The following steps occurred during the Old French period:
Nasalization of /a/, /e/, /o/ before /n/ or /m/ (originally, in all circumstances, including when a vowel followed).
Nasalization occurs before, and blocks, the changes /ei/ > /oi/ and /ou/ > /eu/. However, the sequence /ɔ̃i/ occurs because /oi/ has more than one origin, e.g. 'coin' "corner" < cŭneum. The sequences /iẽn/ or /iẽm/, and /uẽn/ or /uẽm/, also occur, but the last two occur in only one word each, in each case alternating with a non-diphthongized variant: 'om' or 'uem' (ModF 'on'), and 'bon' or 'buen' (ModF 'bon'). The version without the diphthong apparently arose in unstressed environments and is the only one that survived.
Lowering of /ẽ/ and /ɛ̃/ to /ã/; but unaffected in the sequences /jẽ/ and /ẽj/ (e.g. 'bien', 'plein'). The merging of /ẽ/ and /ã/ probably occurred during the 11th or early 12th century, and did not affect Old Norman or Anglo-Norman.
Nasalization of /i/, /u/, /y/ before /n/ or /m/.
The following steps occurred during the Middle French period:
Lowering of /ũ/ > /õ/ > /ɔ̃/. (Note that most /ũ/ come from original /õ/, as original /u/ became /y/.)
Denasalization of vowels before /n/ or /m/ followed by a vowel or semi-vowel. (Note that examples like 'femme' /fam/ "woman" < OF /fãmə/ < fēmina and 'donne' /dɔn/ "(he) gives" < OF /dũnə/ < dōnat, with lowering and lack of diphthongization before a nasal even when a vowel followed, prove that nasalization originally operated in all environments.)
Deletion of /n/ or /m/ after remaining nasal vowels (i.e. when not protected by a following vowel or semi-vowel). Hence 'dent' /dɑ̃/ "tooth" < [*/dãt/] < OFr 'dent' /dãnt/ < EOFr [*/dɛ̃nt/] < dĕntem.
The following steps occurred during the Modern French period:
/ĩ/ > /ẽ/ > /ɛ̃/ > [æ̃]. This also affects diphthongs such as /ĩẽ/ > /jẽ/ > /jɛ̃/, e.g. 'bien' /bjɛ̃/ "well" < bĕne; /ỹĩ/ > /ɥĩ/ > /ɥɛ̃/, e.g. 'juin' /ʒɥɛ̃/ "June" < jūnium; /õĩ/ > /wẽ/ > /wɛ̃/, e.g. 'coin' /kwɛ̃/ "corner" < cŭneum. Note also /ãĩ/ > /ɛ̃/, e.g. 'pain' /pɛ̃/ "bread" < panem; /ẽĩ/ > /ɛ̃/, e.g. 'plein' /plɛ̃/ "full (m.s.)" < plēnum.
/ã/ > /ɑ̃/.
/ỹ/ > /œ̃/. In the 20th century, this sound has low functional load and has tended to merge with /ɛ̃/.
This leaves only four nasal vowels /ɛ̃/, /ɑ̃/, /ɔ̃/, and /œ̃/, and increasingly only the three /ɛ̃/, /ɑ̃/, /ɔ̃/.
フランス語の基底言語と外来言語のインパクト
French is noticeably different from most other Romance languages. Some of the changes have been attributed to substrate influence—i.e. to carry-over effects from Gaulish (Celtic) or superstrate—influence from Frankish (Germanic). In practice, it is difficult to say with confidence which sound and grammar changes were due to substrate and superstrate influences, since many of the changes in French have parallels in other Romance languages, or are changes commonly undergone by many languages in the process of development. However, the following are likely candidates.
In phonology:
The reintroduction of the consonant /h/ at the beginning of a word is due to Frankish influence, and mostly occurs in words borrowed from Germanic. This sound no longer exists in Standard Modern French (—it survives dialectally, particularly in the regions of Normandy, Picardy and Wallonia); however a Germanic h usually disallows liaison: les halles/lɛ.al/, les haies/lɛ.ɛ/, les haltes/lɛ.alt/, whereas a Latin h allows liaison: les herbes/lɛzɛrb/, les hôtels/lɛzotɛl/.
The reintroduction of /w/ in Northern Norman, Picard, Walloon, Champenois, Bourguignon and Bas-Lorrain[15] is due to Germanic influence. All Romance languages have borrowed Germanic words containing /w/, but all languages south of the isogloss —including the ancestor of Modern French ("Central French")—converted this to /ɡw/ (which remains in some words like e.g. linguistique), which usually developed subsequently into /ɡ/. English borrowed words both from Norman French (1066 – c. 1200 AD) and Standard French (c. 1200–1400 AD), which sometimes results in doublets such as warranty and guarantee.
The occurrence of an extremely strong stress accent, leading to loss of unstressed vowels and extensive modification of stressed vowels (diphthongisation), is likely to be due to Frankish influence, and possibly to Celtic influence, as both languages had a strong initial stress. (e.g. tela -> TEla -> toile)[16] This feature also no longer exists in Modern French. However, its influence remains in the uniform final word stress in Modern French—due to the strong stress, all vowels following the stress were ultimately lost.
Nasalisation resulting from compensatory vowel lengthening in stressed syllables due to Germanic stress accent
The development of front-rounded vowels /y/, /ø/, and /œ/ may be due to Germanic influence, as few Romance languages outside of French have such vowels.
The lenition of intervocalic consonants (see above) may be due to Celtic influence: A similar change happened in Celtic languages at about the same time, and the demarcation between Romance dialects with and without this change (the La Spezia-Rimini Line) corresponds closely to the limit of Celtic settlement in ancient Rome. The lenition also affected later words borrowed from Germanic (e.g. haïr < hadir < *hatjan; flan < *fladon; (cor)royer < *(ga)rēdan; etc.), suggesting that the tendency persisted for some time after it was introduced.
The devoicing of word final voiced consonants in Old French is due to Germanic influence (e.g. grant/grande, blont/blonde, bastart/bastarde).
In other areas:
The development of verb-second syntax in Old French (where the verb must come in second position in a sentence, regardless of whether the subject precedes or follows) is probably due to Germanic influence.
The first person plural ending -ons (Old French -omes, -umes) is likely derived from the Frankish termination -ōmês, -umês (vs. Latin -āmus, -ēmus, -imus, and -īmus; cf. OHG -ōmēs, -umēs).[17]
The use of the letter k in Old French, which was replaced by c and qu during the Renaissance, was due to Germanic influence. Typically, k was not used in written Latin and other Romance languages. Similarly, use of w and y was also diminished.
The impersonal pronoun on "one, you, they" – (from Old French (h)om, a reduced form of homme "man") is a calque of the Germanic impersonal pronoun man "one, you, they", reduced form of mann "man" (cf Old English man "one, you, they", from mann "man"; German man "one, you, they" vs. Mann "man").
The expanded use of avoir "to have" over the more customary use of tenir "to have, hold" seen in other Romance languages is likely to be due to influence from the Germanic word for "have", which has a similar form (cf. Frankish *habēn, Gothic haban, Old Norse hafa, English have).
The increased use of auxiliary verbal tenses, especially passé composé, is probably due to Germanic influence. Unknown in Classical Latin, the passé composé begins to appear in Old French in the early 13th century after the Germanic and the Viking invasions. Its construction is identical to the one seen in all other Germanic languages at that time and before: «verb "be" (être) + past participle» when there is movement, indication of state, or change of condition; and «"have" (avoir) + past participle» for all other verbs. Passé composé is not universal to the Romance language family—only Romance languages known to have Germanic superstrata display this type of construction, and in varying degrees (those nearest to Germanic areas show constructions most similar to those seen in Germanic). Italian, Spanish and Catalan are other Romance languages employing this type of compound verbal tense.
The heightened frequency of si ("so") in Old French correlates to Old High Germanso and thanne
The tendency in Old French to use adverbs to complete the meaning of a verb, as in lever sus ("raise up"), monter amont ("mount up"), aler avec ("go along/go with"), traire avant ("draw forward"), etc. is likely to be of Germanic origin
The lack of a future tense in conditional clauses is likely due to Germanic influence.
The reintroduction of a vigesimal system of counting by increments of 20 (e.g. soixante-dix "70" lit. "sixty-ten"; quatre-vingts "80" lit. "four-twenties"; quatre-vingt-dix "90" lit. "four-twenty-ten") is due to North Germanic influence, first appearing in Normandy, in northern France. From there, it spread south after the formation of the French Republic, replacing the typical Romance forms still used today in Belgian and Swiss French. The current vigesimal system was introduced by the Vikings and adopted by the Normans who popularised its use (cf Danish tresindstyve, literally 2 times 30, or 60; English four score and seven for 87)[要出典]. Pre-Roman Celtic languages in Gaul also made use of a vigesimal system, but this system largely vanished early in French linguistic history or became severely marginalised in its range. The Nordic vigesimal system may possibly derive ultimately from the Celtic. Old French also had treis vingts, cinq vingts. (cf. Welsh ugain "20", deugain "40", pedwar ugain "80" lit. "four-twenties").
Holmes Jr.,Urban T.;A. H. Schutz(1938).A history of the French language.Biblo & Tannen Publishers.pp.29.ISBN0819601918
ホームズ・Jr.,アーバン・T、アレキサンダー・H・シュッツ著、松原秀一 訳『フランス語の歴史』大修館書店、1974年、42頁。ASINB000J94F0I。
Eric Hobsbawm, Nations and Nationalism since 1780: programme, myth, reality (Cambridge Univ. Press, 1990; ISBN 0-521-43961-2) chapter II "The popular protonationalism", pp.80–81 French edition (Gallimard, 1992). According to Hobsbawm, the main source for this subject is Ferdinand Brunot (ed.), Histoire de la langue française, Paris, 1927–1943, 13 volumes, in particular volume IX. He also refers to Michel de Certeau, Dominique Julia, Judith Revel, Une politique de la langue: la Révolution française et les patois: l'enquête de l'abbé Grégoire, Paris, 1975. For the problem of the transformation of a minority official language into a widespread national language during and after the French Revolution, see Renée Balibar, L'Institution du français: essai sur le co-linguisme des Carolingiens à la République, Paris, 1985 (also Le co-linguisme, PUF, Que sais-je?, 1994, but out of print) ("The Institution of the French language: essay on colinguism from the Carolingian to the Republic. Finally, Hobsbawm refers to Renée Balibar and Dominique Laporte, Le Français national: politique et pratique de la langue nationale sous la Révolution, Paris, 1974.