Geography in Psalms: Hagiosophitikon Settings of Psalms 1, 2 and 3

The term hagiosophitikon is quite well-known in Byzantine chant, where it occurs in manuscripts from the 14th century onwards: Most often one encounters the designation hagiosophitikon in rubrics of settings of Psalm 1 (Μακάριος ἀνήρ) and Psalm 3 (Κύριε, τί ἐπληθύνθησαν οἱ θλίβοντές με). Nevertheless, neither the exact meaning of hagiosophitikon has ever been clearly determined, nor have the settings themselves been melodically analysed yet. 
Thus, hagiosophitikon is sometimes explained to denote chants associated with services in the church of Hagia Sophia in Constantinople and in Thessaloniki respectively, and/or to represent hymns rather conservative in outline or to have been composed in a very local style. 
The present article will aim at both clarifying the term’s meaning itself as well as presenting a detailed analysis of the compositional style by comparing the hagiosophitikon settings of Psalms 1 and 3 among themselves and with melodies named palaion (old) in manuscripts of the 14th and 15th centuries: What kind of compositional style do the hagiosophitikon settings display, do they use formulas and if yes which ones, might their syllabic parts show traces of the old, so-called “simple” psalmody and how are the melismatic parts treated? – These are just some of the questions which will be discussed in the article in order to get closer to resolving the unanswered questions surrounding the term hagiosophitikon.

Some other designations inserted in the rubrics are e.g., boulgarikon ("Bulgarian"), dysikon ("Western"), frangikon ("Frankish"), persikon ("Persian"), hagioreitikon (from Mount Athos), thessalonikaion (from Thessaloniki) etc. 2 , 1 "The Development of a Didactic Tradition. The Elements of the Papadike," in Tradition and Innovation in Late-and Postbyzantine Liturgical Chant 2: Proceedings of the congress held at Hernen Castle, the Netherlands, 30 October-3 November 2008, eds. G. Wolfram-Chr. Troelsgård (Leuven: Peeters, 2013 See also the list in the index in Diane H. Touliatos-Miles, A Descriptive Catalogue of the Musical Manuscript Collection of the National Library of Greece. Byzantine Chant and Other Music Repertory Recovered which -as Flora Kritikou has already shown -either point to the origin of the chant or indicate a certain influence from the geographical locations mentioned. 3 These geographical designations seem to go hand in hand with another development in fourteenth-century codices, namely, the inclusion in the rubrics for a given chant of the name of the composer to whom that chant is attributed. 4 Hagiosophitikon is also one of those terms that is mentioned only cursorily in footnotes and margins of studies in Byzantine chant. Neither has its exact meaning ever been clearly determined, nor have the settings carrying the designation been melodically analysed. Therefore, we cannot tell for sure if this geographical designation simply points to the origin of a chant or also to a certain way of chanting it.
To make things even more complicated, it is not clear -either from the manuscripts themselves or in the secondary literature -which Hagia Sophia-church is indicated by hagiosophitikon. In her catalogue of music manuscripts in the National Library of Greece, Diane Touliatos claims in one instance that hagiosophitikon chants are associated with the Hagia Sophia in Constantinople and in another with the Hagia Sophia in Thessaloniki respectively. 5 Edward Williams is careful not to commit himself, when he writes in his book on Ioannes Koukouzeles about the first three psalms: "Present for the first time among the anonymous works in Athens 2458 are several settings accompanied by the rubric 'Hagiosophitikon', which may represent chants associated with services in the Great Church of Hagia Sophia in Constantinople." 6 The question of the precise referent of the hagiosophitikon chants has for some time been unanswerable; unless other sources turn up, providing more clues, it can only remain guesswork as to which Hagia Sophia was meant. We shall return to this question below.

Manuscript sources and chants
Chants called hagiosophitikon cannot be found in a great number of codices. After carefully studying all well-known relevant and accessible manuscript collections, we can compile a list of approximately thirteen manuscripts from the early fourteenth until the late fifteenth centuries which I have chosen as the deadline for the present article. 7 GR-An 2458, the earliest (Aldershot: Ashgate, 2010), 631f. 3 Cf. Flora N. Kritikou, "Byzantine Compositions entitled 'Dysikon' (Western) and 'Fragikon' (Frankish): A Working Hypothesis on Potential Convergence Points of two Different Traditions," Journal of the International Society for Orthodox Church Music 3 (2018): 191: "In Byzantine musical manuscripts a number of compositions entitled thetalikon, politikon or persikon are regularly found. As is generally accepted, titles as thetalikon or politikon indicate an analogous origin for these chants, while, respectively, in the case of persikon an influence from a so-called 'external chant' is suggested. In the same way, other titles as dysikon and fragikon, meaning 'Frankish' and 'Western', are also detected; these, according to the practice of Byzantines scribes and composers, denote a western or Frankish origin and/or a certain influence of western music and liturgical practice, respectively." 4 Cf. Edward V. Williams, "John Koukouzeles' Reform of Byzantine Chanting for Great Vespers in the Fourteenth Century" (PhD diss., Yale University, 1968), 214. 5 Touliatos-Miles, Catalogue, 474 and 395. 6 Williams, "John Koukouzeles '", 214. 7 Of course, other codices similar to those included in the list also contain hagiosophitikon settings. manuscript of the Akolouthiai-(or Order of Service), is also the earliest codex to include an hagiosophitikon setting. Older manuscripts, such as the heirmologia ET-MSsc 1256 and 1257, from 1309 and 1332 8 respectively, already display psalm compositions but do not contain chants with the designation hagiosophitikon. This might be due to the fact that these early codices do not attribute any psalm settings to composers, but contain only anonymous compositions. Online scans: https://bit.ly/3DpqkKB. 10 No online scans available. 11 Online scans: https://bit.ly/3dq294k. 12 The hagiosophitikon chant cannot be found among the kratemata settings from fol. 35r onwards as claimed by Touliatos,Catalogue,377. 13 Online scans: https://bit.ly/3GeGIPW. 14 Online scans: https://bit.ly/3EoW4AS. 15 Online scans: https://bit.ly/3dosOyh. 16 Online scans: https://bit.ly/3ptUcRg. 17 Online scans: https://bit.ly/3dph4M2. 18 Online scans: https://bit.ly/31vyXpM. 19 Online scans: https://bit.ly/3EybOBF. 20 Online scans: https://bit.ly/3Gfj8Te. 21 Online scans: https://bit.ly/3pv1GDr. 22 Online scans: https://bit.ly/2ZYvEqN.
The term hagiosophitikon seems to be applied almost exclusively to verses of Psalm 1 (Μακάριος ἀνήρ) and to a lesser degree to those of Psalms 2 (Ἵνα τί ἐφρύαξαν ἔθνη) and 3 (Κύριε, τί ἐπληθύνθησαν οἱ θλίβοντές με). So far, I have only found one other psalm and three non-psalmic chants that are called hagiosophitikon: These settings differ radically from melodies of psalms 1-3 termed hagiosophitikon. In these cases the denomination hagiosophitikon is definitely faulty and seems to have been caused by errors on the part of the copyists (please see part "Wrong/incorrect attributions" for detailed analyses).

Structural and stylistic analysis
Looking at the compositional style of the chants in question puts one on more stable ground than mere assumptions as to the meaning of the term hagiosophitikon itself. Edward Williams 24 offers an interesting hint when he calls the hagiosophitikon settings "conservative". He claims that the "[…] relatively conservative 'Hagiosophitikon' chants have lost this near balance between length of Psalm text and length of Alleluia, for the music of the refrain is more than twice the length of the Psalm text." Williams goes on to show 25 that the hagiosophitikon chants comprise an average of 35 notes for the psalm text and 76 for the Alleluia, thus doubling the notes for the refrain. Unfortunately, Williams does not disclose the way in which he counted the neumes: It is not clear, for instance, if he counts two combined neumes such as Apostrophos+Elaphron ( ) as two notes or just as one. Therefore, I come to a different ratio regarding the length of the verse and the refrain in hagiosophitikon settings: Based on my own transcriptions (see below), the Alleluia refrain usually exceeds the psalm verses by approximately ten notes (I count neume combinations that are sung as one note only as one and not as two).
Contrary to the psalm verses, the Alleluia refrains are freely composed settings that do not follow any discernible pattern. Here, the composers seem to have been "allowed" to use their own creativity more than when dealing with the psalm-verses themselves where -apparently -they had to adhere to older, traditional formulas, handed down from generation to generation. The present article will therefore concentrate on the melodic analysis of the verses themselves and not on the Alleluia refrains. On account of the very different style of the refrains, they still await detailed analysis to explore the possibility that they too might manifest a different, though still distinct, hagiosophitikon compositional style. 26 The highly formulaic verses, on the contrary, will provide more insights into the oral past of psalm-verses and thus also possible connections to the so-called "simple" psalmody. The term "simple psalmody" was coined by Christian Troelsgård who describes it is "flexible and orally administered type chanting" which "seems to be quite stable and firmly linked to the eight-modes-system" which "formed the musical 'backbone' of the Byzantine office". 27 Taking a close look at the structure of the verses themselves it becomes apparent that they consist of two parts: a) Incipit: A strictly syllabic beginning which resembles a kind of recitation, rather than a proper melody, and which in its basic outline is common to Psalms 1, 2 and 3. This recitation is made up of tone repetitions (isa ) and no intervals larger than an ascending or descending second for which only petasthai ( ) and apostrophoi ( ) are used, thus making it easy to determine where the melisma starts: 26 The lively responses to my latest papers on various aspects of Byzantine music has shown that interest is especially high regarding chants with geographical designations. The present article can therefore be regarded as the starting point for further research. Among others, the author of this article will examine chants termed thettalikon/thessalonikaion in a paper to be presented at the 8 th International Conference of the RASMB-IMS Musical Cultures and Diasporas in the Balkans (Aug/Sept 2023 in Thessaloniki). Concerning the complex questions of the Alleluia refrains, the author will submit a paper dedicated exclusively to this topic at Leeds International Medieval Congress in July 2023. Furthermore, the definite aim of this article is to inspire future studies on both Alleluia refrains of various Byzantine chants as well as on other chants with geographic denominations which have hitherto remained unexamined. 27 On the concept of "simple psalmody" see above  This simple recitation is thus highly adaptable to the different length and syllable counts of the various verses as well as to the text accents. As becomes apparent when comparing the incipit of the hagiosophitikon chants, the melodic line on an accented syllable always goes up a second, usually using a petasthe. 28 This type of incipit, however, is not unique to hagiosophitikon settings, as it is also used by composers in chants termed palaion and anonymous ones (see e.g. A-Wn Theol. gr. 185, fol. 9v for the anonymous setting of Psalm 1, verse 3c καὶ τὸ φύλλον αὐτοῦ or fol. 10r, Psalm 1, 5b οὐδὲ ἁμαρτωλοί). Thus, the incipit gives us an idea of how psalms might have been sung before they started to be embellished: They probably consisted only of these syllabic recitations with a short formula at the end. This is what the so-called "simple" psalmody must have looked like (see also Figure 4 for simple psalmody). 28 My thanks go to Charles M. Atkinson for pointing out this feature, which is an important characteristic of Western psalm-tones, whose cadences are expanded or contracted to reflect accented syllables. b) Melodic part: The melodic part starts right after the shorter or longer syllabic incipit and consists strictly speaking of a melisma on a chosen syllable, usually the fourth syllable from the end of the verse. Comparing the chosen verses of Psalms 1 to 3 it becomes apparent that there are two distinct melodies for these melismata (I call them A and B) plus one variant (A' and B') each. These melodies are quite easy to recognize, making the hagiosophitikon chants clearly discernable for the listener and probably making them easy to remember for the singers. The red rectangle in Figure 3 shows the syllabic incipit discussed above. The blue circles contain melody A with the transcription into Western staff notation. This melody is used for all the verses shown in Table 3: As one can see in Figure 3, the biggest interval here is an ascending fourth at the beginning of the melisma; otherwise there are only ascending or descending seconds and thirds; the ambitus covers a sixth. The melisma always appears on the fourth syllable from the end of the verse, except for verse 6 of Psalm 3 (ἐγὼ ἐκοιμήθην καὶ ὕπνωσα), where it is on the third syllable form the end. Usually, the melody starts on the syllable before the melisma. Melody A also has a short melisma on the final syllable of each verse in a distinct cadential formula (see the green oval in Figure 3). If, for experimental reasons, we take away the melisma on ἀποῤῥυήσεται (i.e. Melody A) we probably obtain the old syllabic (simple) version:  Melody A' is a kind of shorter variant of Melody A; regarding for instance verse 6b of Psalm 1 (καὶ ὁδὸς ἀσεβῶν ἀπολεῖται) in Figure 5, this appears as follows: the red box exhibits the common syllabic beginning, the blue circles the melody for the small melisma on τῶν which is a shortened version of Melody A (the transcription into Western staff notation in Figure 5 below gives an idea of the melodic outline of the melisma).

Figure 5: Outline of Melody A' (above) in comparison to Melody A (below)
GR-An 2458, fol. 15v (Psalm 1, verse 6b: Καὶ ὁδὸς ἀσεβῶν ἀπολεῖται) (© Athens National Library) GR-An 2444, fol. 26v (Psalm 1, verse 3c: Kαὶ τὸ φύλλον αὐτοῦ οὐκ ἀποῤῥυήσεται) (© Athens National Library) The melody shown in the blue oval in the first example of Figure 5 (taken from GR-An 2458, fol. 15v) corresponds with that in the blue oval of Gr-An 2444, fol. 26v. It is a shortened version of the melisma that starts with an ascending fourth ( ) in Gr-An 2444 -a leap that is omitted in the shortened version.
GR-An 2401 and GR-An 899 also label their settings of the last word of verse 6b (ἀπολεῖται) hagiosophitikon (see Figure 6 below): GR-An 2401 might provide a glimpse into the syllabic setting of this verse because it uses one formula from Melody A shown in Figure 3 above without the short melisma on ἀπολεῖται, employing the four neumes found on ἀπολεῖται for the whole word, which is a very common simple cadence.    Figure 7 shows again the melody of the melisma in blue circles: Like Melody A, Melody B also employs a fourth (descending this time) as the largest interval, but otherwise it consists again of ascending and descending seconds and thirds. The ambitus of Melody B is an octave, and thus slightly larger than the range of Melody A. I could not find any proof for Williams's claim 29 that "[…] the total spectrum of range reveals that the 10th is by far the most popular vocal ambitus for the 'Hagiosophitikon' […] chants of the first Stasis 30 […]." Contrary to Melody A, Melody B has no common final formula; all the melismata end differently here before the Alleluia refrain begins. A variant of Melody B appears at the end of verse of 13b 2 of Psalm 2 (οἱ πεποιθότες ἐπ᾿ αὐτῷ), taking as its melodic substance only parts of melody B, as indicated by the blue circles in Figure 8 below. 29 Williams, "John Koukouzeles'", 235. 30 A stasis is one of three sections of each kathisma, i.e. the twentieth part of the Psalter. Although Melody A is the predominant one for Psalm 1 and Melody B for Psalm 2 (one cannot point to a characteristic melody for Psalm 3 because only very few hagiosophitikon chants can be found for this text), it becomes apparent that these melodies can be "borrowed" by any of the other psalms: Melody A, for instance, is also taken over for one verse of Psalm 3 and Melody A' for Psalm 2. Melody B can be found for a verse of Psalm 1 as well as for Psalm 3. Therefore, I cannot concur with Williams, "John Koukouzeles'", 31 who claims that "inter-psalm" migration, where "melodies might wander from one psalm to another or appear in all three psalms of the first Stasis […] does not occur in any of the anonymous settings, i.e. the quasi-traditional or Hagiosophitikon, but is only a feature of certain newly-composed melodies." As shown above, the two main Melodies A and B do occur in all three psalms of the first kathisma.

Wrong/incorrect attributions
Occasionally, settings are designated hagiosophitikon in the rubrics of manuscripts, although they do not seem to belong to this type: These seemingly incorrect designations are spotted quite easily, once one knows the incipit and the melodies typical for hagiosophitikon chants (see the Figures above): The chants so designated either contain longer melismata with unusually high pitches and/or larger intervals or use a different incipit and a melody that is different from the standard ones for hagiosophitikon settings. However, one must be aware of the possibility that the application of the term hagiosophitikon to chants other than the verses of Psalm 1-3 could mean something else stylistically, or simply be a non-stylistic term or use or origin: Musicians of this period might have used this term in a broader sense to denote a chant somehow related to the usages of one or more churches called Hagia Sophia. Bearing this in mind, I found three verses of Psalm 2 and one of Psalm 3 (see Table 7 above) which show uncharacteristic features for hagiosophitikon chants: • GR-An 2406, fol. 42r, for instance, does not begin with the characteristic incipit but with a fifth upwards. The short melody displayed afterwards does not match the typical hagiosophitikon melodies either. That verse 11b of Psalm 2 (καὶ ἀγαλλιᾶσθε αὐτῷ ἐν τρόμῳ) is termed an hagiosophitikon in GR-An 2406 (see Figure 9 below) is due to a scribal error: 1) because while the verse starts with the common incipit, the melisma on αὐτῷ is much longer with unusual high pitches and large intervals (fourths and fifths) than hagiosophitikon chant; 2) this is confirmed by a comparison with a setting found in A-Wn Theol. gr. 185: On fol. 16v of this manuscript it can be seen that the chant found in GR-An 2406, fol. 43r is a kalophonic composition attributed to Christophoros Mystakon(os) (mid-fourteenth century 32 ) and not a hagiosophitikon (see  A-Wn Theol. gr. 185, fol. 16v (Psalm 2, verse 11b: καὶ ἀγαλλιᾶσθε αὐτῷ ἐν τρόμῳ; setting with attribution to Mystakonos -see the transcription above) (© Austrian National Library) • In GR-An 2444, fol. 32r (Psalm 2, verse 13b 2 : οἱ πεποιθότες ἐπ᾿ αὐτῷ) the incorrect designation also seems to be due to a scribal error: Αs the ending of this half-verse is preceded by the whole verse itself (μακάριοι πάντες οἱ πεποιθότες ἐπ᾿ αὐτῷ), which constitutes a hagiosophitikon setting (see Table 4), the scribe probably thought the following repetition of the words οἱ πεποιθότες ἐπ᾿ αὐτῷ was also a hagiosophitikon. Furthermore, the setting on fol. 32r does not begin with the characteristic incipit, but only with pitch repetitions and its melisma is much longer, containing ascending and descending fifths. • In GR-An 2458, fol. 19v, verse 6b of Psalm 3 (ἐξηγέρθην, ὅτι Κύριος ἀντιλήψεταί μου) is named an hagiosophitikon in the rubric. The verse does start with the common incipit, but follows with an unusual melody and a melisma much too long for a hagiosophitikon chant. That the hagiosophitikon incipit is used here is not uncommon, as it is also taken up frequently by composers for their own settings of psalm verses (see e.g., the melody by Xenos Korones of verse 6b of Psalm 3 in GR-An 2444, fol. 33r).

Conclusion
To sum up the findings regarding the chants called hagiosophitikon: In most cases by a wide margin the term hagiosophitikon is ascribed to settings of Psalms 1, 2 and 3. Such geographical attributions seem to be much less random than has heretofore been assumed, and they are attached to specific psalms/ hymns (e.g., in the hesperinos prokeimena in A-Wn Theol. gr. 185 from fol. 66r onwards, many instances of politikon and thettalikon/thessalonikaion appear). Svetlana Kujumdzieva 33 also states for the kekragaria that "some sources add the designations 'politikon' […] and 'Thessaloniki' to it." What prompted the wish to ascribe psalm-settings to special areas or to remember where they came from? In addressing this question I can concur with Dimitris Balageorgos 34 who states that there were probably two different chant traditions in existence during the fourteenth century, a situation that created the necessity of reforming the terminology so as to provide not only composers' names and specific attributions regarding the age of individual chants (i.e. palaion/old and neon/new), but also their style (e.g. organikon/ instrumental) and their geographic connections, as cited at the beginning of this article. Another reason can be found in the extensive production of new chants that reached an unprecedented and unheard-of peak in fourteenth century-Byzantium, which might have caused the scribes/singers to feel it necessary to facilitate a differentiation between the compositions.
Regarding melodic style and range, the hagiosophitikon chants can be said to preserve an older tradition, probably even older than the settings called palaion (old), which tend to be longer and more embellished. The hagiosophitikon chants are shorter in both their verses and their Alleluia refrains than are the palaion settings and those attributed to specific composers. Thus, Williams is definitely right when he calls the hagiosophitikon settings "conservative" in outline.
This article has shown that the hagiosophitikon chants provide hitherto unknown clues for traces of the so-called simple psalmody: As Psalms 1 to 3 show, the melodic formulas employed for hagiosophitikon chants were easy for singers to remember and could also be adapted to various verses regardless of their length and syllable count.
Regarding the uncertainty as to which church could have been meant by "in the style of the Hagia Sophia", it seems safe to assume that the one in Constantinople was the intended referent. The hagiosophitikon chants are quite widely spread and appear in a greater number of manuscripts than has hitherto been assumed, which makes it plausible that they were developed in the great church of the capital, rather than in the smaller one in Thessaloniki, which is itself an emulation of the Hagia Sophia in Constantinople. From there, these easily recognized, remembered, and chanted settings started their "journey" through the realm of Byzantium and their inclusion in many of the fourteenth/fifteenth-century manuscripts. This assumption is also confirmed by settings in the Polyeleos (Psalms 134, 135, 136) called "Voulgara" or "Voulgarikon," where the melodies making up the melismata shown above were reminiscent of a melodic practice at the Hagia Sophia. 35 Furthermore, there exists the geographic denomination thessalonikaion/ Thessalonian or "in the style of Thessaloniki" which points to chants from that city, so that we can safely assume that hagiosophitikon was attributed to settings from the Hagia Sophia in Constantinople.