Scansion

Prosodic analysis into metric quantitative

The metric diagrams are abstract: they are not specific of a given language. The same diagram (like, for example that of the dactylic Hexameter) can be useful in theory as well in Greek as in Latin, or in any language who is ready to use it. One can represent a metric diagram like a sequence of “boxes empty”, large or small, called positions and intended to accommodate syllables: syllables of worms given, in a given language. A metric diagram also evokes a musical line, made up of white and the black ones, but without words. One remembers that there exists, for the metric positions, the two categories heavy (-), “white”, and the light (U), the “black ones”.

By definition, one understands here by Prosodie the part of the grammar which treats accent and quantity in a given language, without particular regard to versification (with the study of the worms and their structure, one holds the denomination of Métrique ). The concrete application of a metric diagram abstracted to worms given in a particular language rests on the prosody of this language. It is indeed the prosodic analysis which, basically, makes it possible to locate the quantity (or length) “natural” of the syllables taken individually and to determine which syllable can occupy which metric position. More still than its Phonology, the prosody of the dead languages poses extremely thorny, and probably insoluble problems. However, the little that one knows some is enough for the reader eager in general to establish the scansion of a poem.

The prosody of Latin, like that of the Greek, knows the two categories long and in short for which one tends to use the same signs (- and U) that into metric. This “homonymy” can cause confusion. As it will be seen, there exist indeed cases where syllables which one would note U under the terms of the prosody must with the final being stressed -, and vice versa. An additional difficulty comes owing to the fact that the prosody regards as two different properties, though partially bound, quantity of the vowels and that of the syllables, one not being always identical to the other for a given syllable.

Syllabication

Posted metric constituting the diagrams being intended to receive syllables, it is necessary to initially examine the syllabication of the traditional languages, which rests on simple conventions. One will cut out in syllable without taking account of the words. With this intention, one can simply consider that the worms constitutes one length word then to cut out it in syllables (attention not to cut out a diphthong, true or distorts, in two vowels):

Tityre, you patulæ recubans sub tegmine fagi. (Virgile, Bucolique I, towards 1)

→ Tityretupatulærecubanssubtegminefagi.

→ Ti.ty.re.tu.pa.tu.læ.re.cu.bans.sub.teg.mi.ne.fa.gi.

Χρυσέῳ ἀνὰ σκήπτρῳ, καὶ λίσσετο πάντας Ἀχαιούς. (Homère, Iliade , I, towards 15).

→ Χρυσέῳἀνὰσκήπτρῳκαὶλίσσετοπάντασἀχαιούς.

→ Χρυ.σέ.ῳ.ἀ.νὰσ.κήπ.τρῳ.καὶ.λίσ.σε.το.πάν.τα.σἀ.χαι.ούς

The point marking separations of syllables, one will retain these examples that:

each syllable counts only one vowel or a diphthong;
only one consonant according to, in the same word or not, a vowel belongs to the following syllable ( Ti.ty.re and not *Tit.yr.e , χρυ.σέ.ῳ and not *χρυσ.έ.ῳ);
Except exception, of two consonants (or more) according to a vowel, the first belongs to the same syllable as this vowel, the second (and the following ones) with the following syllable ( teg.mi.ne and not *te.gmi.ne or *tegm.i.ne , λίσ.σε.το and not *λίσσ.ε.το or *λί.σσε.το). The case of consonants followed by liquids ( L and R ), even the nasal ones ( N and m ), is treated further.

It is advisable to distinguish the open syllables (i.e. whose core or vocalic nucleus is followed of no consonant or coded) closed syllables (whose nucleus is followed of one coded). For example, the Latin word armed cuts out in ar.ma . The first syllable, rear , is closed: its nucleus, have , are followed by one coded, R . Second is open: there is nothing after its nucleus, has .

Quantity of the syllables

In certain cases, it is possible to know the quantity of a syllable without being concerned with that of its vowel. In other cases, that one results from this one.

Short syllable

A syllable is short containing only one short vowel in an open or followed syllable only one consonant :

λόγος = λό.γος = UU (λό- is a syllable containing a short vowel in an open syllable; - γος contains a short vowel followed by only one consonant);
in the same way in Latin: tămĕn = tă.mĕn = UU.

To identify a short syllable requires necessarily ─ but that is not sufficient ─ that it is known that its vowel is short. No general rule not making it possible to determine it, this knowledge rests on a detailed study of the prosodic quantity.

Long syllable

A syllable is long containing:

a long vowel in open syllable. One speaks then about length by nature :

δώρῳ = δώ+ρῳ = - - (ω is a long vowel),
in Latin: ā+mās = - -,

a vowel, it does not matter its prosodic quantity, followed by two consonants (at least) marked (that these consonants belong or not to the same word). In this case, the syllable is known as long by position :

ἐξ = - → although ε is a short vowel, it is followed of two marked consonants; in καλὸν δαιδάλεον the syllable - λὸν is stressed - because ο, although prosodiquement short, is followed of two consonants,
in Latin: rex = -; in Jovis pax , the syllable - screw is stressed - although the I is prosodiquement short here.

To identify a long syllable, when it is opened or that its vowel is followed only one consonant, requires that it be known that it is prosodiquement long

Double letters

One will take guard with the following fact well: the Greek and Latin use single signs noting two consonants (they are named sometimes double letters ), which thus lengthen automatically the syllable:

Greek: ζ = then at the time hellenistic (but not, value that this letter forever A), ξ =, ψ =;
Latin: J (also written I ) =, X =, Z = (these two last letters are borrowed from the Greek).

Note: for ζ / Z and J , one could just as easily note and, are respectively “/z/ length” and “/j/ length”.

Thus, one stresses:

πεζός = - U, ἄξω = - -, ἔρριψα = - - U;
ejus = - U, pax = -, Gaza = - U.

The Greek aspired consonants (θ =, φ =, χ =), however, are not to regard as double letters. Moreover, the initial aspiration, materialized by the hard Spirit, does not count for a consonant: in λόγον ὑπέρϐαλε, the syllable - γον is stressed U whereas, phonologiquement, one has /l ógo' N h' upérbale/.

Placement of the two consonants

If the two consonants belong to the following syllable, the syllable is not lengthened. One can symbolize that as follows:

VoyelleConsonne+C → long syllable by position;
V+CC → long or short open syllable according to the length of the vowel.

For example, the syllable - gĕ in light stator is stressed U. This rule however valid, is not thus formulated, that in Latin. For the Greek, it is enough that two consonants follow a vowel, whatever the place of the consonants, so that the syllable is lengthened: the syllable - γᾰ in μέγα στύγος is stressed -.

One will see in the paragraph according to whether during time a particular case developed, however.

Groups with liquid (and nasal for the Ionian-attic)

What precedes rises directly from the preceding rule.

One names traditionally liquid the consonants /r/ and /l/. Those follow a particular treatment in the scansion. Indeed, when they are preceded by another consonant ( B , C , D , F , G , p and T for Latin, β, γ, δ, θ, κ, π, τ, φ and χ for the Greek), they inevitably do not form a group of two consonants likely to lengthen the preceding syllable.

Thus, one can to consider that patris stresses pat.ris = - U or pa.tris = UU, which makes say to certain theorists who the syllable Pa is prosodiquement common , i.e. at will long or short. In the same way in Greek: πατρός is worth πατ.ρός = - U or πα.τρός = UU. It should be noted that this second treatment (known as correptio attica “correption attic”) is foreign with the language of Homère and meets practically only in Ionian-attic, which extends it from the remainder to the consonant groups + nasal (μ or ν). The Homeric language is unaware of it thus and one always stresses πατ.ρός = - U (one notes some poetic licenses, however, allowing to avoid impossible rates/rhythms in the worms).

In kind, that confirms that the lengthening of the syllable is made only if one of the two consonants already belongs to the syllable to be lengthened, which is then closed: V+CC does not lengthen, contrary to VC+C.

The /w/ consonants and qu

In Greek
In Greek, the Phoneme /w/ (of wadding /wat/), inherited the Indo-European , is quickly amuï. Although still used in Mycénien, it disappears as of the first millenium before the Christian era, except in certain dialects. The Greek of Homère being a composite language, some of the words employed coming from Ionian preserve old a /w/, which is however not written, consonant that the scansion makes it possible to reveal to explain lengthenings which, if not, would seem irregular. Moreover, the Homeric forms with /w/ are confirmed by the etymology (they are thus not poetic licenses). For historical reasons, one notes the consonant by means of the letter Digamma capital ─ Ϝ in , ϝ into tiny (to note that it is of this letter that comes the F Latin) ─ used in certain versions of the Greek alphabet to return the sounds or (in the remainder of the text, one will make use only of the capital, more often included in the font faces).

The /w/ phoneme being a consonant, it can thus play in the quantity of the syllables:

lengthening by position:

digamma at the beginning of word: ὃς εἴδει… → must be read ὃς Ϝείδει, which makes it possible to stress - - - and not U - - (one syllabifie ὃσ.Ϝεί.δει and not ὃ.σεἴ.δει),
digamma in a word: ἐκερέω → ἐκϜερέω = - UU - and not UUU - (ἐκ.Ϝε.ρέ.ω and not ἐ.κε.ρέ.ω),

prevention of the correption (shortening of a long vowel or a diphthong at the end of the word in hiatus; to see low): ἀλλ' ἤτοι ἔπεσιν → ἀλλ' ἤτοι Ϝέπεσιν = - - - UU and not - - UUU;
prevention of the elision: κατὰ ἄστυ → κατὰ Ϝάστυ (one would wait if not κατ' ἄστυ).

Latin
The /w/ consonant is normally noted by means of the same sign as the /u/ vowel. It is only in the modern editions that they are distinguished ( v = and U =; they are the “Lettres ramists”). This notation, however, is not without ambiguities: indeed, if v is always a consonant, U can also the being:

in the Digraph qu : U does not constitute a Phonème but the labiovelar appendix of the phoneme complexes /k ʷ/. qu ─ always in front of vowel ─ thus counts only for only one consonant. Ăquă stresses UU: qu only cannot lengthen a syllable;
Cu in front of vowel is worth never /k ʷ/but /kw/, two consonants, which can thus lengthen the preceding syllable;
gu in front of vowel does not form a digraph (it does not exist phoneme /g ʷ/compared to /k ʷ/. However, in front of vowel, gu is sometimes carried out /gw/, which makes two consonants, sometimes /gu/ (consonant + vowel). It is not possible to slice in a peremptory way. For example, in lingua , one will read /lin.gwa/ = - - but arguet will be worth /ar.gu.et/ = - UU;
in front of U in front of vowel and preceded by other consonants can be read /w/: suadeo /swa.de.o/.

Quantity of the vowels

To determine that a syllable is long is easier than for the short syllables. Indeed, these last cannot be regarded as such as if one is sure quantity of the vowel whereas for the long syllables the presence of two consonants according to a vowel, if necessary, is enough.

Each vowel of the Greek or Latin (but also of the Sanskrit and many other languages) has an intrinsic quantity that the practice, the dictionary and the C-W communication can help to determine: mălŭs , “malicious”, mālŭs , “apple”, ἔργον/ĕrgŏn/, “work”, θρῦλος /thr ūlŏs/, “rumor”, etc This quantity intrinsic goes, directly or not, to make it possible to determine the scansion.

Short vowels

In Greek

The Greek alphabet is less ambiguous than the Latin : indeed, the noted vowels ε and ο are always short (in opposition to η and ω, always long). The letters α, ι and υ remain ambiguous: they can as well represent ᾰ, ῐ and ῠ that ᾱ, ῑ and ῡ. Thus, a word as ἔλεγον can be directly stressed UUU.

In the worms epic (the dactylic Hexameter), it can happen that a long vowel at the end of the word in Hiatus with the initial vowel of a following word shortens. This phenomenon cannot of course occur if the long vowel is on the key period of the foot. One names that the correption . For example, ἄνδρα μοι (ἔννεπε) will be stressed like a Dactyle - UU and not - U - because - οι is shortened by correption in front of the ἐ- according to. This correption would not be possible if μοι constituted the first time of the foot because this one, in a hexameter dactylic, is necessarily long. Thus, μοι ἐπὶ would be well stressed - UU.

In Latin

The Latin alphabet, although not making it possible to distinguish the quantities (at least such as one currently writes it), another index offers: indeed, in this language (but not in Greek), a vowel in Hiatus (followed by another vowel) in the same word is shortened ─ vocalis handle vocalem corripitur ─ (even if a H separates them because this letter is dumb): the I in the following words can thus be identified like briefs: nĭhil , omnĭa , filĭus . This rule suffers some exceptions however:

the E radical of the 5th variation remains long between two I : diēi but rěi ;
in the pronominal variation, the Ending of Genitive - ius generally stresses - īus ;
in the Form S without - R of the verb fieri , the I radical is long: fīat but fĭerem .

In addition, a final vowel followed by only one consonant other than - S is short: legĭt , amĕm , tamĕn , sorŏr . Followed S , the quantity of such a vowel cannot be given (except for - ās and - ōs , always long). In the monosyllables, this rule does not function systematically, however.

Long vowels

In Greek

The vowels η and ω are always long.

The Diphtongue S are long, even the false : αι, ει, οι, αυ, ευ and ου. The diphthongs with first long element are not considered longer still: ᾳ (i.e. “ᾱι”), ῃ (= “ηι”), ῳ (“ωι”) and ηυ. Thus, the Iota subscribes being able to be only under one long vowel, a ᾳ can be only long.

Any vowel resulting from a Contraction is long: - α of the imperative τίμα is thus long since the Forme goes back to *τίμαε.

Finally, the Ending S - οι and - αι ─ which is considered short (except with the optative and in certain adverbs) are to be stressed -. Thus, λῦσαι /l ūsăi̯/is stressed - -.

As only the long vowels can carry the circumflex accent, this one makes it possible to identify long when it is about ᾶ, ῖ and ῦ (ῆ and ῶ being already obviously long by the only presence of η and ω).

Let us recall that the long vowels can undergo the correption.

In Latin

The diphthongs are always long: æ , œ , with the and have .

Finally, - ace and - bones always have a long vowel: - ās and - ōs .

The vowels resulting, by Monophtongaison, of a diphthong remain long. Certain monophtongaisons can be detected only by studying the history of the language. Others are however easily visible: they are those resulting from the Apophonie. To know the radical of a verb makes it possible to determine that a monophtongaison could occur:

on the radical of cæd-o , one obtains the perfect with redoubling EC-cīd-I . It is obvious that the /i/ of the radical coming from/æ/, it remains long;
the radical claud-o makes it possible to obtain derivatives in - clūd , from which the central vowel is resulting from /au/: -clūd-O , idiot-clūd-O , in-clūd-O , Re-clūd-O , ex-clūd-o , etc

In the same way, the vowels resulting from a contraction are necessarily long. The contraction is however a very frequent Latin phenomenon who asks, once again, a certain historical knowledge. Certain cases are however easy to retain:

Génitif contract. The genitive of the names of the second variation in - ius or - ium are normally *ii . One however finds very frequently (and it is in fact the oldest form) a vowel contracts - ī : sestertius → sestertii / sestertī , consilium → consilii / consilī , etc;
Vocatif S contract. Same names (with the masculine) in - ius has a vocative in - ī resulting from a contraction: filius → filī ! ;
perfect and pluperfects contract. In the conjugation, parvaits in - avi , - avisti , etc or - ivi , - ivisti , etc, and pluperfects in - veram , - will veras , etc, which result from this can be “syncopated” (by Amuïssement of intervocalic /v/ then possible contraction of the vowels in contact). The vowels while resulting are of course long: audivisti → audīsti , amavisti → amāsti , deleveram → delēram , etc;
other contractions to be known: nihil → nīl , *ne (H) emo (i.e. homo ) → nēmo , mihi → mī , etc

Elisions

The vowels in hiatus between two words are generally elided. They thus do not count any more in measurement of the worms. The elision S and Aphérèse S (elisions opposite) Greek do not raise any difficulty: indeed, they all are replaced by the apostrophizes: ἀλλά ἐγώ → ἀλλ' ἐγώ, ὦ ἄναξ → ὦ ῎ναξ.

The elisions of Latin, on the other hand, are more awkward since nothing, in the writing, indicates them. However, an elided vowel does not count any more in the worms. The rule is the following one: any vowel (short, long, diphthong) at the end of the word are elided in front of a word begin with a vowel. Two details are notable:

the written consonant H does not decide. It thus does not prevent the elision: vidi homines is stressed as if one wrote vid' homines ;
the - m at the end of the word, as of the traditional time, did not decide any more. It thus does not prevent either the elision: hominum agmen is read hominu agmen , from where homin' agmen .

Synereses

The Synérèse (or synizèse ) is the fact of reading in only one syllable two vowels which do not form normally a diphthong (it is a kind of Contraction by Coalescence in which the two vowels keep their identity). The vowel or the diphthong while resulting is long (but can be shortened by correption). It is a poetic license which it is sometimes necessary to suppose to stress worms correctly. It is thus not possible to give immutable rules as for its employment.

Here are some examples (for the scansion of the worms and not of the only syllables, to see low):

Pasco libatis dapidus. Cuic Prout libido is (Horace, Satires 2,6,67).

In it towards, prout must be read in only one long syllable (let us recall that or is not a Latin diphthong), that is to say /prowt/ (or, in API, even). In it towards prout is thus stressed _ instead of U_.

Μῆνιν ἄειδε θεὰ Πηληϊάδεω Ἀχιλῆος (Homère Iliade I, 1).

It is necessary to read the syllable - δεω with a syneresis.

One cannot really know how all synereses were marked: indeed, if prout can easily be carried out, it is not easy to determine how - δεω was said. Was this a diphthong? A long vowel?

To determine the feet and the caesura

Once one determined the quantity of each syllable of which one can deduct the quantity, it is necessary to cut out the towards in feet then to place the Césure. Each meter having his own rules of scansion, one will return to two examples to give an idea of the method to be followed:

Method to stress a hexameter dactylic (see also Hexameter dactylic );
Method to stress trimeter iambic a (see also Trimeter iambic ).

There exist however general rules which must be mentioned here.

Substitutions by contraction and resolution

It was seen, it exists only two metric categories, the heavy category (which would be equivalent for example to white) and the light category (black). However, two black being worth white, it is often possible to replace, in a foot, a heavy position awaited by two light or two light awaited by a door:

contraction: UU → -;
resolution: - → UU.

Possibilities of resolution and contraction depending however on the meter. For example, it is possible, in the dactylic Hexamètre, to transform each Dactyle (- UU) into a Spondée (- -). It is however not true in the Pentamètre S of the elegiac distiches where the last two dactyls cannot be spondées. It is thus necessary to refer to the diagram of each towards to know possible substitutions.

Syllable anceps (undifferentiated)

At the end of any worms, the last position is known as undifferentiated ( syllaba anceps in Latin). That means that it can accommodate any syllable, whether it is prosodiquement long or short. In certain meters, other positions which the last can be anceps .

Interrogations and problems of interpretation: how to say the ancient worms?

To say (or sing) worms is an art, and nothing or almost informs us about the practice of this art in Antiquity. What one knows well today is a “school” scansion whose only claim is, without hardly of regard to historical exactitude, to make apparent the metric diagram. This school scansion rests in theory on the “quantity” of the syllables, but she teaches that syllables as different as mā , măs , măns , mās or māns would have the same property, namely that they would be long . However, reality was perhaps very different: who says to us that, in perception intimates that a Roman speaker lambda could have prosody of his own language, all these syllables was actually perceived like length equivalent? For the Acoustic phonetics, indeed, măns will tend to being longer than măs (there is a consonant moreover) and māns that măns . One should thus consider that there exist various degrees of syllabic quantity, in the order ascending: would long vowel → long vowel + consonant → long vowel + consonant + consonant, etc Of these syllables not very long , rather long , long , very long , extra-long , it be thus the metric one (more than the prosody) which would force equivalence, by offering only one category to accommodate them all?

How those which sang the worms with art managed these tensions? Today, it is not known, for example, if it is necessary to distinguish the pronunciation from a syllable which is long at the same time by nature and by position ( māns ) from a syllable which is long only by position ( măns )… Moreover, one cannot often determine which is the exact support of the syllabic quantity. When a syllable is long by position and that its vowel is short, how to support its length so that, in scansion, it lasts twice as much as one short syllable? In a syllable like ăp. (keep silent) , it will be very difficult to make rest the impression length on coded, made up of occlusive that one will have much sorrow “to hold” as a long time as a vowel. Despite everything the efforts, ăp risk to appear shorter than ā . Is it plausible that, with it only, the absence of Désocclusion which caracérise coded it closed syllables is responsible their “length”?

Such are the questions that it is necessary to be posed since one tries to exceed the stage of the “school” scansion. Often, the “restored” readings that one can hear “cheat”: the reader, indeed, lengthens the short vowels of the long syllables by position… Others, however, bring a different answer and tend to respect more exactly the prosody by not lengthening the short vowels of the long syllables by position, but by prolonging the consonants, so that the auditive impression of duration their is quite due, and not with the only vowels. In this idea, the first syllable in ap.tus will be known as (API), with one without Désocclusion prolonged by a light pause which, in fact, will make the syllable long (but without the support of the voice: a less rigorous pronunciation would return this syllable as follows: ).

Towards:

method to stress a hexameter dactylic,
Method to stress trimeter iambic a,

Prosody;
ancient Literature.

External bonds

Theory

Introduction to Greek Meter.

To listen to the stressed ancient worms

Homeric Singing ─ Year Original Approach to the Performance, how to sing Homère in the most faithful possible pronunciation. One will be able to also find other readings;
Society for the Oral Reading off Greek and Latin literature, site impossible to circumvent offering several Latin extracts and Greeks in restored pronunciation of which some of Stephen Daitz, considered as one of the pioneers of the faithful reading. The extract of the Oiseaux of Aristophane (here) particularly is successful;
Viva Voce, Romance Poetry Recited, recitations of Latin worms in a restored pronunciation.

Random links:

Scansion

Prosodic analysis into metric quantitative

Syllabication

Quantity of the syllables

Short syllable

Long syllable

Double letters

Placement of the two consonants

Groups with liquid (and nasal for the Ionian-attic)

The /w/ consonants and qu

Quantity of the vowels

Short vowels

In Greek

In Latin

Long vowels

In Greek

In Latin

Elisions

Synereses

To determine the feet and the caesura

Substitutions by contraction and resolution

Syllable anceps (undifferentiated)

Interrogations and problems of interpretation: how to say the ancient worms?

Related articles

External bonds

Theory

To listen to the stressed ancient worms