The rhythm and stress in Latin
by: F. Claudius Iulianus
Obviously we have no original record of any native speake of Latin :-) How should Latin sound, like a rapid rhythm similar to that of English? Or with a more uniform distribution of time among its syllabes like in Spanish or in Italian?
I believe evidence supplied by position stress accent in Latin support the hypothesis that rhythm of Latin is not like that of Spanish, nor like the accentual rhythm of English but a moraic rhythm like that of Japanese. I will try to explain this idea:
1) Phonologically the major unit that we find featured in some languages with stress accent is that of stress syntagma or stress group. This unit group together all phonemes around an stress accent. For example in Spanish is formed by two accentual groups [estoj / enkasa].
2) In addition to this unit is the syllabic division, found in almost every language in the world like phonological unit.
3) Some few languages recognize unities (of more than one phonema) minor that syllabes. These unities are named "moras". For example in Japanese a word like = 'Japan' is formed by 2 syllabes but by 3 moras: . This means that this Japanese word is uttered in three steps.
Now, I will intend to show that Latin also distingish internally moras like phonological unities (in fact, if stress rules are considered in terms of moras, they also seem very simple: stress is found in the penultimate vowel anterior to the last vowel).
1) In English, the rhythm of speech function in stress groups: that is, each stress group occupies a similar time in utterance. In fact United States of America is uttered in two steps [junairdsteit / evamerÔkŽ] (this type of rhythm explain satisfactorily reduction or neutralization of some wowels in unstressed position!)
2) In Spanish the rhythm of speech is mainly syllabic: each syllabe take aproximately the same time (irrelevantly if stressed or unstressed!). [This is the reason that make difficult English for Italians and Spanish, that hear it like somethin incomplete]
3) In Japanese (and I think in Latin) the rhythm is moraic, each mora take aproximately the same time: takes three "time unities". If you see at the Japanese syllabaries (kana) you see that the Japanese script is no longer syllabic, but moraic [Mistakenly many books of text say that this script is syllabic]. is written with 3 symbols not 2!
Why I affirm that rhrythm in Latin is moraic? By the evidence supplied by accent!
Ancient grammaticians distingished between short ando long syllabes and formulated rules for stress in terms of this distinction: "stress falls in the penultimate syllabe if this is a short syllabe and in the antepenultimate if it is a long one".
If we redefine these rules in terms of moras, the rule is simpler: "Stress falls always in the penultimate mora anteceding last vowel (conditionlas are removed from this rule!)"
Phonological syllabe have an optional onset, a nucelus and an optional cauda, onset + nucleus constitue a mora, and cauda by itself constitues another mora (This is just the division in Japanese!). Also in Japanese an Latin a long vowel like [a:], [e:] . is formed by two moras = /aa/, /ee/, . therefore enlongement of the vowel is considered as a cauda!
If a syllabe is ended with long vowel or consonant it has 2 moras, these syllabes of two moras are precisely the "long syllabes" of the ancient grammaticians. And the "shor syllabes" of ancient grammaticians in termos of moras have longitude = 1.
Some examples (including position of stress)
do'minus = do'.mi.nus
ami':cus = a.mi'.i.cus
inve'ntus = in.ve'.n.tus
a'rbor = a'.r.bor
Obiously a word of only 2 moras should have stress in the next mora preceding last vowel:
ro'sa = ro'.sa
The rhythm of Latin: these rules seemps to incate that stress in Latin follow moraic rules (like in Japanese), and by generalization, I think the rhytym of Latin is also moraic (that is the fundamental unit of time is the mora).
Naturally this is only a hypothesis, but I think this hypothesis, explain the facts related to stress in a more simple form that is rival. And By virtue of the Okham's razor, is preferible as explication of the facts, given the absence of other empirical evidence :-)