Egyptian metrics - a morphological approach

Presentation and video podcast of the seminar, May 11, 2017 - University of Pisa

published on 12th May 2017, 05:54 pm

Metrics is one of the most fascinating branches of a culture; much more when that culture is represented by an extinct language or by an extinct stage of a language. It is like a box filled with endless treasures for the historical phonologist, for the musician, for the philologist and also simply for the historian of literature. Philologists can fill up lacunae thanks to it; musicians can guess at least the basic rules of a music which cannot be heard anymore; historical phonologists can learn from it some data which could never be understood from prose. Just to show the most common example, if it was not for metrics, we would never imagine that, in classic Latin, any ending containing an /m/ preceded by a vowel is never pronounced before a word beginning with vowel. This does not appear from prosastic texts, for Romans do not use hyphens and, whenever they write a word, they write it entire, regardless of its particular pronunciation in that particular phonotactic context.

Knowledge of metrics is much more meaningful for a stage of a language, like middle Egyptian, of which only consonants have been preserved by Hieroglyphic and Hieratic scripts. Should we know the laws of Egyptian metrics, we could also have a more precise idea of how was a particular word vocalized - a datum which is hidden by the original Egyptian writing and which is only partially desumable from Cuneiform transcriptions, Greek adaptations and loanwords, and Coptic language. Of course, metrics would not tell us the timbre of the vowel (unless some kind of rhyme is documented to have existed, but it seems that this is not the case - at least, not systematically), but it would tell us whether there was a vowel between two consonants or not, and whether this vowel was long (if Egyptian metrics was mora-timed, as I think) / stressed (if it was stress-timed) or not. On the other hand, knowledge of metrics is much more difficult to achieve in Egyptian; and this is just for the same reason: Egyptian writing is vowelless. So, as it is easy to understand, exploring Egyptian metrics is difficult for the same reason why it is important! Because of these problems, however, metrics is one of the less studied fields of ancient Egyptian philology and linguistics. One counts about twenty papers entirely devoted to it, plus some tens of indirectly related ones.

There are three kinds of metrics: stress-timed, syllable-timed and mora-timed. In stress-timed metrics, verses have a constant number of stresses; in syllable-timed ones there are constant amounts of syllables; and in mora-timed ones it is the mora which governs the metre, and which shows constants amounts. Examples are easier to understand than explanations: a sentence like "send it to me" is of two units for a stress-timed speaker: [sénd-it][to-mé], of four units for a syllable-timed one: [sen][dit][to][me] and of six units for a mora-timed one: [se][n][di][t][to][me]. A word like "metrics" is of one unit for the first one, [métrics], two for the second one, [me][trics], three for the third one, [me][tri][cs], and so on. Now, three are the most important Authors who proposed an original theory on Egyptian metrics (Fecht 1964, Foster 1980, Mathieu 1988, '90, '94, '97), but all these three theories are substantially syntax- and prosody-based and for this reason they share the feature of viewing Egyptian metrics as a stress-timed one (stress is a matter of syntax more than of word): unfortunately, you cannot have a precise idea of the prosody in a language which you cannot hear anymore! Hence, you cannot guess how much stresses were in a sentence and according to which syntactic rules. The writing preserves the syntax, but it cannot preserve the sound: the pitchs, the timbres, the dynamics are totally lost. For this reason, the approach I am proposing with this work of mine is, as I defined it myself, 'morphologic', i.e. based on morphology: in other words, based on the structure of the word and not on the structure of the sentence and its prosody.

Egyptian is an Afro-Asiatic language. A.-A. languages are introflexed; this means that they contain lemmata and broken morphs. A broken morph is a discontinuous one. For example, in English, the word "film" is singular: the plural is "films". The -s of "films" is a sound morph. It begins and ends in itself, without interruptions; while the lemma, "film-", is not invaded in its inner part by morphological elements. The same word in Arabic, "film" (which is obviously a loanword from English), has the plural: "'aflàam". This means that, in Arabic, the lemma is not "film", but "flm", i.e., only the consonants bear the meaning; while the morphology is assigned to the vowels: [i] for the singular, "film", ['a]+['a:] for the plural, "'aflàam". In other words, the singular is of the type RiRR, the plural is of the type 'aRRàaR. "R" stands for "radical". The consonants of the lemmata, i.e. of the roots, are called "radicals". Now, [?a]+['a:] is a broken morph, because it is interrupted by a part of the root ([f]+[l]). It is discontinuous. In general, we speak of introflected languages every time a language assigns semantics to consonants and morphology to vowels: introflective phenomena are present in our I.-E. languages too: couples like "foot vs feet", "speed vs sped", "take vs took", and the like, are good English examples of introflection. "Ft", "spd", "tk" are the roots, "oo", "ee", "e", and "a-e", are the morphs. 

What has all this to do with metrics? There is an important fact which should be kept in mind: some radicals are weak. This means that they fall away when they are intervocalic. And what happens when they fall away? It happens that the two short vowels, which were formerly separated by the weak radical, merge, and become one long vowel. For example, the root for "to be" in Arabic is "kwn": as a semivowel, [w] is typically a weak radical. So, the 3rd sing. pers. of the perfect, which should have been *['kawana], undergoes the falling of the intervocalic weak radical and the two short [a] become one long [a]: ['ka:na]. Another example, this time in I.-E., is provided by Greek: the plural of "polis", i.e. "city", is, as everyone knows, "poleis": this is due to the fact that the classic form of the root is "plj": the plural, hence, should have been *['polejes]; but the [j] fell away and the two [e] became one long [e], which in Greek is written with the digram 'ei'. Now I think you can see what has all this to do with metrics.

Long story short: the number of strong radicals tells you the number of the syllables a sentence contains; the number of all the radicals tells you the number of morae a sentence contains. Simply, if we have a series of sentences in which only the strong ones are constant in number, we have a syllable-timed metrics; if the number of all the radicals, weak and strong ones, is constant, the metrics is mora-timed. If both are not constant, the metrics is stress-timed. Let us take as an example an imaginary Arabic sentence like "kaana maata raama", "he was (root: kwn), he died (root: mwt), he desired (root: rwm)": the number of strong radicals (i.e. those we see: kn-mt-rm) tells us the number of syllables: six. The number of radicals (in general, strong and weak ones) (kwn-mwt-rwm) tells us the number of morae: nine. If the following verse is, for example, "safara kataba", "he traveled (root: sfr), he wrote (root: ktb), then the number is constant (six) only if we take into account the sole strong ones: this means that the metrics is syllable-timed. Otherwise, if the following verse is, for example, "safara kataba qaTala", "he traveled, he wrote, he killed (root: qTl)", then the number is constant only if we take into account all the radicals, and not only the strong ones: kwn - mwt - rwm = 9; sfr - ktb - qTl = 9. This means that the metrics is mora-timed. Finally, if the following verse is "safara kataba basmala", "he traveled, he wrote, he recited the basmalah", then the number is always inconstant, for we have a different number both taking into account only the strong radicals and taking into account them all - hence, the unique left metrics is the stress-timed one; and in fact both verses have three stresses: kàana, màata, ràama - sàfara, kàtaba, bàsmala. I applied this method to lyric passages of some middle Egyptian literary works, and by now the result is that each sentence shows a constant amount of radicals - all radicals, not only the strong ones. This suggests that Egyptian metrics, at least in classic times, was mora-timed. A datum which, if confirmed by my future and further studies (this research is at its first steps!), will add another element of proximity of the Egyptian language with its cognates of the A.-A. family, in the most archaic stages of which vowel length had (and in some cases, like classic Arabic, still has) phonemic and metric value. In this video, I share this research of mine with my colleagues of the University of Pisa and, as I hope, with you ;)

Stefano Vittori

about the author