Dies ist eine Internet-Sonderausgabe des Aufsatzes
„Old Armenian and Caucasian Calendar Systems.
2. Armenian hor̂i and sahmi
von Jost Gippert (1987).
Sie sollte nicht zitiert werden. Zitate sind der Originalausgabe in
„The Annual of The Society for The Study of Caucasia“
1, 1989, 3-12
zu entnehmen.

This is a special internet edition of the article
„Old Armenian and Caucasian Calendar Systems.
2. Armenian hor̂i and sahmi
by Jost Gippert (1987).
It should not be quoted as such. For quotations, please refer to the original edition in
„The Annual of The Society for The Study of Caucasia“
1, 1989, 3-12.

Alle Rechte vorbehalten / All rights reserved:
Jost Gippert, Frankfurt 1999-2001


2. ARMENIAN hor̂i AND sahmi0

Jost Gippert

Freie Universität Berlin

      0.1. A comparison of the Armenian and Georgian calendar must not neglect the names of the second and third months of the Armenian year, hor̂i and sahmi, for these names have for long been etymologized by deriving them from the Georgian cardinal numbers ori and sami meaning “two” and “three.”1 This etymology, tempting as it is, was never discussed with due rigidity in Armenology nor were the arguments that Georgian scholars contributed taken notice of. There are three main questions that have to be dealt with: First, are there any traces of a “numerical monthnaming system” in Georgian tradition;2 second, is it probable that cardinal numbers were used in Georgian to denote the position of a month; third, can hor̂i and sahmi be equated formally with the Georgian ori and sami or their predecessors? All these questions were treated in an extensive article by P’. Ingoroq’va in 1932, the results of which can be summarized as follows.3
      0.2. As for the first question, Ingoroq’va points to the text of the “Conversion of Georgia” which contains several data using “numerical” month names such as “the first month,” “the fourth month.” Although it is the ordinal numbers that appear, Ingoroq’va takes them as indicators of an ancient numerical month-naming system in Georgian, because the text is one of the oldest original texts in that language. As for the second question, Ingoroq’va refers to the Old Georgian way of denoting the days of the month. Here, data using ordinal numbers such as meore dġe and data using cardinal numbers such as ori dġe, both meaning “the second day (of a month),” are met with side by side; according to the author's opinion, the latter mode of expression is the autochthonic one.4 Further on, Ingoroq’va discusses the names of the days of the week. As in Greek, these are built analogous to the Semitic principle by counting from the “Sabbath,” but differently from Greek, the cardinal numbers are used once again; cp. ori šabattay with Greek δευτέρα σαββάτων ‚the second day of the week = Monday.' As for the last question, Ingoroq’va interprets the forms hor̂i and sahmi as representing older *yor̂i and *saymi;
these forms would be identical with the Kartvelian ancestors of Georgian ori and sami to be reconstructed by comparison with its sister languages, Svan, Mingrelian, and Laz. Ingoroq’va's argumentation cannot be maintained like this, however. As will be shown, too many details deserve additional remarks or corrections so that we have to take up the discussion again.
      1. The “Conversion of Georgia,” Mokcevay kartlisay, has come down to us in two redactions preserved in one manuscript each. The older one of them, the so called “Šat’berd” version, has to be dated into the second half of the X century which gives us a terminus ante quem for the genesis of the text.5 The “Conversion” must indeed be called “original” inasmuch as no exact model for the complete text, Greek or Armenian, has been found. It can be shown, however, that it was compiled by using different sources available at that time. The major part of the text is dedicated to the life of St. Nino, the legendary missionary of Georgia.
      1.1. The “Conversion” contains four datings that operate with months denoted by ordinal numbers. They appear in two successive chapters of that part of St. Nino's vita, which according to the text was written dewn by one “Salome from Užarma” after the words of St. Nino herself. The first two datings are of Armenological interest, too, because here, St. Nino tells us how she came from Jerusalem to Armenia together with St. Ripsimē and St. Gaianē. First, the chronological background is sketched:

“Then the Lord took part of Greece, and Constantine the Emperor and his mother and all his court adopted faith in him and confessed Christ in the year 5444 since creation and in (the year) 311 since the resurrection of Christ. And all Greece was converted to Christianity. (And) in the year was the Council of Nicaea, and in the eighth year we fled from Greece.”6

The text continues with the following details:

“Ripsime the queen and Gaiane the nurse and 50 souls, we marched off in the first month, on the fifteen(th day) and came to the territory of Armenia, to the gardens of Trdat‛ the King. There they died in the first month, on the thirt(ieth day), a Friday. But I kept (hidden) in the thorns of the roses, for rose and almond were blooming at that time.”

After seeing an angel's vision, St. Nino leaves Armenia:

“And I went off and came to Ulup’oreta, and I spent the winter with great sufferings. And in the fourth month I went to the mountains of J̌avaxeti in order to know. where Mcxeta was.”7

      The fourth dating in question is given some pages later, after St. Nino reports how she observed the Georgian king Mirian worshipping his (Zoroastrian) god Armaz8 and how the idols were destroyed by a thunderstorm that she had caused by praying:

“And when the storm calmed down, I went off .. And there was a beautiful acacia tree, high and with many twigs. I went under that tree and I marked it with the sign of Christ's cross. And I prayed there for six days. When you came up, a crowd of people, and looked for the gods that brought forth suffering, and you didn't find them, then I was there, for it was the sixth day of the (sixth) month, when Evmanoel showed the Father's image to the heads of the living and to the heads of the dead.”9

      1.2. As the Georgian scholar K’. K’ek’elije pointed out, the datings as presented above admit of a coherent interpretation which emanates from the details contained in the last passage mentioned. Here, the text is obviously referring to the miracle of the “Lord's Transfiguration” revealed in the New Testament.10 As this miracle led to the institution of a Christian festival on August 6th, we may assume that this date is meant by the “sixth day of the sixth month.” Thus, we arrive at the month of March to be equated with the “first month” of the “Conversion.” By the same reckoning, the “fourth month” converges with the month of June. This interpretation agrees with other information contained in the text such as the rose and almond blooming at the end of the “first month.” It is further supported by some evidence from outside the “Conversion:” There are some later, “metaphrastic” versions of St. Nino's legend, the most important of which is the one ascribed to the XI century bishop Leont’i Mroveli; it forms part of the great Georgian chronicle, Kartlis cxovrebay.11 Here, the first two datings are omitted; the “fourth month,” however, is glossed by the phrase “which is June.”12 Thus, a Georgian year beginning with the month of March and denoting the months by ordinal numbers seems in fact to have existed.13
      1.2.1. There is another testimony, however, which renders this conclusion rather doubtful. This is the Armenian translation of the Georgian chronicle, Patmowt‛iwn Vrac‛. As the oldest manuscripts of the Georgian text have to be dated as late as the XV century, the Armenian translation, which was probably accomplished some three centuries before, is a witness of utmost importance for establishing the original wording. As for the legend of St. Nino, the Georgian and the Armenian versions diverge to a considerable extent. The latter does not reflect the “metaphrastic” vita of Leont’i Mroveli bot an older text standing closer to the “Conversion” inasmuch as it is told in the first person.14 If we now compare the passages in question we are struck by the fact that it contains but one of the datings, viz., the third one, and that the month is not denoted by the ordinal number here:

Ew im yarowc‛eal eki yOwrbanis hayoc‛, ew jmereal and, ew i yownis amsean eki i lear̂n J̌awaxet‛ic‛ “and I got up and went to the (town of the) Urbans (in the land) of the Armenians, and I spent the winter there, and in the month (of) June I went to the hills of J̌avaxeti.”

The attestation of the “Transfiguration Day” is rendered as follows:

Ew im .. eki ar̂ car̂ovn br̂nč‛i .. ew aławt‛ec‛i and zvec‛ awr. Ew yawowr meci Gownap‛oxowt‛ean Tear̂n, yoržam ec‛oyc‛ Tērn zpatkern hawr glxaworac‛ ar̂ak‛eloc‛ ew margarēic‛n .. “And I .. went to the acacia tree .. and I prayed there for six days. And on the great day of the Lord's Transfiguration, when the Lord showed (his) Father's image to the main apostles and prophets ...”15

These details suggest, that the version standing behind the Armenian text is anterior to both Leont’i Mroveli's and the one of the “Conversion” and that the latter also bears some additional adornments.16 And we have good reasons to believe that just the “numerical month-naming system” contained in the “Conversion” is due to such additions. Taking the model that has to be assumed for the Armenian text as the primary source, we can suppose the following development: At first, only one month dating occurred in St. Nino's report, viz., the month of “June.” This would match other parts of St. Nino's legend as contained in the “Conversion,” where the Latin month names of July, March, and May appear.17 The other datings, howover, originated in marginal notes, which later penetrated into the text.
      1.2.2. This assumption can be motivated as follows. According to the Armenian tradition manifested in Agathangelos's history, St. Ripsimē and St. Gaianē came to death on the days of 26 and 27 hor̂i, resp. These dates were rendered as 26 and 27 September in the Greek version of the “History,” and the Old Georgian translation gives 26 and 27 stulisay.18 There has been, however, a secondary arrangement in the Greek ecclesiastical calendar by which the holy women together with St. Gregory the Illuminator became associated with the 30th of September.19 This association spread into the Georgian tradition, too, as can be seen by the fact that the Old Georgian version of the Martyrology of St. Ripsimē and St. Gaianē is headed in the manuscripts with the date of ttuesa sek’denbersa L ,“in the month of September, 30(th).”20 As September was the first month of the Byzantine “indictional” year, it is this date of a “30th of the first (‚a') month” which might have been added as a marginal gloss to the model of St. Nino's legend as contained in the “Conversion.” Thus, the “first month” can only secondarily have been equated with the month of “March.” And indeed, it is only the “Conversion” which speaks of “rose and almond blooming at that time;” cf., however, the Armenian Patmowt’iwn: Ew mnac‛i es i nerk‛oy vardeneac‛, or oč ēr całkeal “And I kept (hidden) amidst the rose trees, which were not blooming.”21 The assumption that the martyrdom of St. Ripsimē and her companions should have taken place in an autumn month rather than the month of March, further agrees with the fact mentioned in all versions that St. Nino after hiding in the roses went to the town of Urbnisi and spent the winter there.
      1.2.3. A marginal gloss of different character can be responsible for the “fifteenth of the first month” being given as the date of the Saints' flight only in the Šat’berd text of the “Conversion.” If we presume that the “first month” is due to an adaption to the date of the “martyrdom” (= September), the number “fifteen” may have been introduced as an emendation of the “eighth year after the conversion of Greece.” There is indeed a contradiction in the datings noted in the text: The adoption of Christianity by Constantine the Great might well have been in 311 AD; the Council of Nicaea, however, did not take place seven years but fourteen years after this event, viz., in 325. It was for this reason that the later (Č’eliši) variant of the “Conversion” changed the year of 311 into 318.22 As the Saints' flight is said to have happened one year after the Council of Nicaea, we exactly arrive at a space of fifteen years counting from 311.
      1.2.4. As for the mention of the “Transfiguration Day” on the “sixth day of the sixth month,” we first have to note that it is only the Č’eliši version which explicitely states the number of the month, the Šat’berd text speaking of the “sixth day of the month” only.23 If we further compare the “metaphrasis” of Leont’i
Mroveli, we are led to the assumption that the “sixth month” is due to a perseveration of the number “six:” Here, we are even told that the event took place in the “sixth hour” of the “sixth day and this month.”24 And it is the Armenian Patmowt‛iwn again which
suggests that the original text contained the number just once, viz., in the statement that St. Nino “prayed for six days.” In any case, a later marginal gloss naming the “6 August” as the date of the “Transfiguration Day” is likely to have penetrated into the text in a corrupted form to give the attested indications.25 And it is equally probable, now, that the equation of the month of “June” with the “fourth month” became necessary as an adjustment after the “first month” and the “sixth month” had been introduced into the text.
      1.3. Thus, Ingoroq’va's and K’ek’elije's assumption of an Old Georgian “numerical month-naming system” revealing itself in the “Conversion of Georgia” can hardly be maintained. In particular, the attestations do not allow for the conclusion that the method of denoting months by (ordinal) numbers was an archaic feature of Georgian. They cannot be used for arguing in favor of the hypothesis that the Armenian hor̂i and sahmi should have been borrowed from Karlvelian stock.
      2.1. As long as no authentic usage of numbers for denoting the months of the year has been attested in Old Georgian, we cannot decide with any certainty, of course, whether cardinals or ordinals would have been used for this purpose in any Kartvelian language loaning hor̂i and sahmi to Armenian. In this connection, we are indeed left alone with Ingoroq’va's arguments cited above, which referred to the mode of naming the days of the week and of expressing the day of a month in Old Georgian. As for the latter feature, however, Ingoroq’va's opinion that the usage of cardinal numbers is the autochthonic one cannot be taken for granted. If we take as examples the attestations of the old month names within hagiographical literature, datings operating with ordinals and datings using cardinals are met with in nearly equal quantities.26 There is a preponderance, however, according to the value of the number in question: Ordinals seem to be preferred with numbers below ten, cardinals with higher numbers.27 Before trying to evaluate this remarkable fact, we have to consider that nearly all the hagiographic texts must be traced back to Greek models. Bearing this in mind, the disproportion can be motivated by the following suggestions: In Greek, it was the ordinal numbers that were commonly used. They were not written explicitely in every case, though; with higher numbers, the “symbolic” writing using letters such as κα' for “21” was preferred. The translators who endeavored to keep near to the original wording retained the ordinals as they were. Wherever they came upon a “symbolic” denotation, they were free, however, to use their own mode of expression. Thus, Ingoroq’va's assumption that the cardinals reflect the proper Georgian diction has a good deal in its favor.28 But, of course, it is not a decisive argument for hor̂i and sahmi being loans from Kartvelian.
      2.2. Lastly, the hypothesis is problematical with regard to the sound equivalences involved. If we want to identify the Armenian month names hor̂i and sahmi with the Georgian numerals ori and sami, we have to cope with the most significant difference between them, which consists in the sound h occurring in the Armenian forms only. As Old Georgian possessed a h- sound of its own, there is no reason why the Georgian numerals should not contain this sound, too, if they really were the models for the Armenian names. In this connection, we have to deal with the proposal made by Ingoroq’va, again, that the borrowing had taken place in ProtoKartvelian times. Ingoroq’va's assumptions match the communis opinio inasmuch as the number “two” must be reconstructed as *yori.29 Some caution is necessary, however, in the case of “three,” which he reconstructs as *saymi. The attested forms of this numeral are sami in Georgian, sumi in Mingrelian and Laz, and semi in Svan. Here, we have the crucious correspondance of Georgian a and Svan e, which is opposed to the more regular correspondance of Georgian a and Svan a or a, reflecting a ProtoKartvelian *a. As the conditions that led to a Svan e in cases like semi have not yet been stated with certainty,30 Ingoroq’va's reconstruction of a word internal “diphthong” ay can be regarded as a possible solution of the problem.31 There would even be some evidence in favor of this solution if Ingoroq’va's assumptions made for the Armenian hor̂i and sahmi would prove correct. It is hardly tenable, however. that these forms replace older *yor̂i and *saymi. In the former case, this would presuppose that the Middle Armenian change of y- to h- should have manifested itself in a word occurring as early as, e.g., in Agathangelos's history.32 In the case of sahmi, attested in Agathangelos as well,33 Ingoroq’va assumes a development of -ay- to -ah- for which no parallel whatever has thus far been attested. ln other words, we should expect the Armenian names to appear as *yor̂i and *saymi if they were really borrowed from the Proto-Kartvelian forms as proposed by Ingoroq’va. As long as such attestations are wanting, any attempt of deriving hor̂i and sahmi from Georgian or from a
previous stage of this language remains undemonstrated.


Achtung: Dieser Text ist mit Unicode / UTF8 kodiert. Um die in ihm erscheinenden Sonderzeichen auf Bildschirm und Drucker sichtbar zu machen, muß ein Font installiert sein, der Unicode abdeckt wie z.B. der TITUS-Font Titus Cyberbit Unicode. Attention: This text is encoded using Unicode / UTF8. The special characters as contained in it can only be displayed and printed by installing a font that covers Unicode such as the TITUS font Titus Cyberbit Unicode.

Copyright Jost Gippert, Frankfurt a/M 13. 8.2001. No parts of this document may be republished in any form without prior permission by the copyright holder.