(BEING CONTINUED FROM 2/08/16)
It has been generally accepted that Glagolitic was patterned on Cursive Greek (Taylor 1881, Leskien 1905, Jagić 1911, Vajs 1932). This is not only testified to by historical manuscripts, such as Xrabr’s “On Letters,” but also supported by the structural similarity between the Glagolitic and Greek alphabet systems.10
Greek has 24 letters in a fixed order, and each letter bears a name, phonetic value, and numerical value. The Greek alphabet consists of three groups, each of which represents 1-9, 10-90, and 100-900. Since there are only 24 letters, the system includes three symbols that only represent numbers (i.e., numerals). Glagolitic is identical to Greek in that each letter has a name, phonetic value, and numerical value, appearing in a fixed order. Glagolitic has 38 letters and thus can represent more numerical values than Greek does: 1-9, 10-90, 100-900, and 1000-9000 (according to Mathiesen, forthcoming, 9). For representing numerical values, Glagolitic uses 36 individual letters and two digraphs (ⰑⰫ [u/ü], ⰟⰋ [y]). Each individual letter’s name, order, numerical value, and phonetic value in Greek and Glagolitic are given in Table 1.
In Table 1, we can observe that the numerical values of Greek and Glagolitic letters are not identical because the letters for Slavic-only sounds intervene. It is noteworthy that the relative order of letters and their numerical values are the same in Glagolitic and Greek with only a few exceptions, which shows that Constantine first arranged Slavic sounds according to the order of corresponding Greek letters, and then inserted Slavic-only sounds where relevant in terms of the class of sounds.12
Although it is clear that Constantine used the Greek alphabet as his primary model in creating Glagolitic, the directory of sounds in Greek was not identical with that in Slavic: some Greek sounds did not exist in Slavic, while some Slavic phonemes were not represented among Greek letters. It is not difficult to imagine that Constantine had to resolve these problems in different ways. As Mathiesen (forthcoming, 23-24) describes, for the twenty-one sounds in common, Constantine adopted Greek letters (I will set aside the issue of formal similarity/difference between corresponding Greek and Glagolitic letters until Section 3.2). For the Greek-only sounds, he evaluated whether each one would be necessary for translating the Bible and preaching to Slavic people. As a result of these evaluations, five letters, Ⱛ (=υ), Ф (=φ), Ⱈ (=χ), Ⱉ (=ω ), Ⱇ (=θ), were added to Glagolitic. Finally, it was crucial for Constantine to create symbols to represent the ten Slavic-only phonemes. Letters for these sounds must have been modeled on scripts, other than Greek, as will be explored in detail in the next section.
It is widely accepted among Slavists that the letter for the sound /š/ was adopted from Semitic. However, Cubberley (1984) and Mathiesen (forthcoming) instead focus on the phonetic similarities of /š/, as well as some other sounds, between Glagolotic and Armenian.13 Mathiesen compares seven Glagolitic letters that represent non-Greek sounds with seven Armenian ones and observes that the relative order of five of these letters is identical, as illustrated in Table 2.
Mathiesen advances a claim that Constantine, after exhausting Greek letters, turned to the Armenian alphabet to represent the remaining Slavic sounds. Given that Constantine’s mentor Bishop Photius was particularly well-versed in Armenian, it is very likely that Constantine had at least partial knowledge of Armenian and that he paid attention to the Armenian alphabet system when he was looking for models while creating Glagolitic. Mathiesen points out that Armenian is included in the list of various older languages that Constantine enumerated in his discussion with Roman church scholars (VC chapter XVI). Although Mathiesen cannot explain the other three letters Ⱏ [ǝ/ĭ], Ⱑ [ě], Ⱒ [x] in the same way, he still maintains their correlation with Armenian, pointing out that Armenian also has distinct phonemes of Խ [x], Ը [ǝ], and Է [ē] (the last of which is different from Ե [e]).
Although Constantine adopted Greek as a structural model for Glagolitic, it does not necessarily follow that he also took Greek as a model for the shapes of the letters, especially given the apparent formal discrepancies between Greek and Glagolitic.14 In terms of letter shapes, Glagolitic looks like a patchwork consisting of shapes from various sources. However, it is unlikely that Constantine used models randomly from various script systems. The formal similarities alone are not a reliable basis for suggesting a correlation between scripts. In addition to physical resemblance, the availability or accessibility of the scripts, on the one hand, and the religious justification for adopting certain models according to the historical context, on the other, should also be taken into consideration.
(TO BE CONTINUED)
Seoul National University, KOREA
10 Xrabr’s account further describes how Constantine-Cyrill was sent by God to the Slavs “to compose 38 letters, some according to the shape of Greek letters, some according to the Slavic word.”
11 Table 1 does not include four pairs of letters that exist in Cyrillic and Greek but not Glagolitic, nor does it include two digraphs. I follow Mathiesen as regards the order of letters, but the details of the ordering do not bear on the discussion in this paper.
12 This similarity is not only found in the comparison between Glagolitic and Greek. Many alphabets that were influenced by Greek appear more or less similar to Greek, assigning a name, a numerical value, and a fixed order to each letter.
13 See also Gaster (1887) and Vernadsky (1943) for the link between Armenian and Glagolitic.