The SONHIS Project

Towards the recovery of the lost SOuNds of HISpanic chant

A comparative study of medieval chant through
computational recognition of neume and melodic patterns

The "Aquitanian" chants in Aemilianensis 56

In the section about the Hispanic neumes (which we recommend to read before what follows) we anticipated that a bunch of chants were able to jump over the notational gap that condemned the vast majority of Hispanic repertoire to silence. Most (16) of these surviving chants belong to the Office for the Dead (Ordo in finem hominis diei) of the Hispanic rite.

It happens that around the times in which the Hispanic rite was in the process of being substituted by the Franco-Roman –at the end of the 11th or the beginning of the 12th century–, the original Hispanic notation for these sixteen pieces was rasured from a Liber ordinum preserved today in the Royal Academy of History in Madrid with catalogue signature 56 -source referred hereafter as Aem. 56 (Aemilianensis no. 56). The erased neumes were then substituted by the Aquitanian dot-notation in which the intervalic quantities can be read -no without difficulties due to the very cramped new writing and the absent in some cases of the tonal reference line typical of the Aquitanian notation [1].

For instance, the five neumes at the beginning of the Educ domine antiphon in Silos 4 we used to ilustrate the uncertainties associated to the Hispanic neumes have now the Aquitanian counterpart in A56 showed in Figure 1.

Figure 1. First five neumes of antiphon Educ domine in Aem. 56, fol. 91r. Neumes have been artificially color-enhanced in blue for better identification. In this case, traces of the rasured Hispanic neumes can be observed -the 3rd and 4th original neumes (a cursive pes and an angled clivis on the syllables do and mi) can even be identified.

In these chants the Aquitanian neumes are referenced to a given final pitch (the finalis note) using both tonal dry-lines and the gregorian differentia annotated at the margin of each chant (see, for instance, [2] and [3] for details). In any case, we can now assemble a duplex edition of these chants in which the pitches encoded in the Aquitanian transcriptions in Aem. 56 are accompanied by the Spanish neumes that for the same chants are found in the Silos 4 source (another Liber ordinum). In particular, Figure 2 shows the duplex score resulting for the initial neumes of the antiphon Educ domine.

E-S ms 4, 091r   


Aem. 56, 028v   










Figure 2. Duplex score for the first five neumes of the antiphon Educ domine. Hispanic neumes after Silos 4, f. 91r; pitches after Aquinatian notation in Aem. 56 f. 28v.

The index below lists the sources in which the 16 pieces transcribed into Aquitanian notation in Aem. 56 can be found. Two preces belonging to the same Office for the Dead are added (no. 3 and 4) which were identified (also written in Aquitanian notation) in a 11th-century Franco-Roman Gradual ("Graduale Albiense", F-Pn lat 776) as presumably imported from any of the older Hispanic sources [4]. At the difference of the highly degree of concordance between the Aquitanian neumes in Aem. 56 and the Hispanic ones in Silos 4 (both belong to the same local Hispanic tradition of La Rioja [5]) there are some mismatches between the Hispanic neumes and the pitch content in the F-Pn lat 776. Without excluding other hyphotheses, as an inmediate adaptation to Franco-Roman melodic tastes [6], such discordances could be due to any or both of the following causes: i) the preces suffered a process of melodic divergence after the tranfer, ii) the transfer was made from a different melodic tradition other than the Riojan one -if we only had the 13 Hispanic correspondences the other 16 Aem 56 "Aquitanian" chants that are found in the Leon Antiphonary (see León 8 entries in the index below), manuscript which defines the Leonese tradition[7], the number of discrepancies between the lower and upper lines of our duplex scores would have increased significantly.

The staff transcription of Aquitanian notation appearing in the lower part of our duplex scores is largely based on the early edition by Rojo y Prado [8], taking also into account the later revisions made by Randel (for the responsorial verses [9]) and by Asensio [10], the latter for the booklet accompaning the record of these chants performed in Solesmes style by Schola Antiqua. Close inspection to the Aem. 56 manuscript led to minor amends. Additionally, for some multi-neume syllables (in which the manuscripts apparently discord) the sounds encoded in the Aquitanian neumes were regrouped into the combination providing the best match with the Hispanic notation. The square notation used in all previous editions has been substituted by the Volpiano format -sometimes, Volpiano conventions about spaces between neumes, syllables or words have been relaxed to accomodate the wider neumes of the upper Hispanic line.

Taken together, these 18 chants should constitute a sort of Rosetta stone providing the key to decipher the meaning of the Hispanic neumes. Additional information can be obtained from manuscripts with origin in the scriptorium of the Silos monastery and now conserved in the British Library (Add MS 30847, Add MS 30848, Add MS 30850), sources that, despite their Franco-Roman content, were written with Hispanic notation shortly after the abolition of the Hispanic rite. The first conclusions obtained by historical studies that contrasted neumes and pitches in these chants and sources point to the fact that, if considered as isolated entities, each form of neume does not involve a univocal relationship with specific pitches or intervals [e.g. 11]. However, one of the founding hypothesis of the SONHIS project is the existence of more semantic levels to explore in the sequences of neumes, levels for which there is already a partial confirmation in the literature of the field.

So we need to start talking about neume patterns.

Chant Index

no.cantus IDincipitgenreE-S ms 4E-Mh cod 56E-S ms 3E-L ms 8GB-Lbl Add MS 30851F-Pn lat 776
1h00100Ecce ego viamRS(B)085r024r015r, 027r*(sm)
2h00020Dies mei transieruntRS(B)087v026r016v277v200r
3h00101Deus misererePRCS088v027r(h)138r
4h00111Miserere misererePRCS090r138r
5h01719Educ domineA091r028v019v236r
6h01720Emitte manum tuamA091r028v019v151v
7h01462Exsurgat deusA091v028v019v(sm)122r026r
8h01730In protectione deiA091v029r021r127v049r
9h00086De manu inferniRS(G)093v030v022r146r
10h01489Credo quia redemptorA094v031v277r
11h01488Memorare domineA095r032r023r(sm)277r
12h01470Si ascendero in caelumA096r, 100v*(sm)32v, 37r*(h)024r(sm)151v
13h01480Manus tua deusA096r033r130r
14h01487In pace in idipsum (x2)A092r, 097r, 100v(sm), 103r(sm)029v(h), 033v, 037r*(h), 038r*(h)021r135v
15h01454Aperiat tibiA097r, 100v(sm)033v, 037r*(h)025v
16h01483Requiem aeternamA097v034r
17h01492Terra terraA098r, 100v*(sm)035r, 037r*(h)025r277r
18h01697Sinite parvulos (x2)A036v, 100r036v028v(sm)092r, 227r*

Genres: A, antiphon; PRCS, preces; RS(B) responsory (with tone B verse); RS(G) responsory (tone G verse).

*, only incipit.

(sm), sine musica (no music, only text).

(h), incipit or duplicated chant with hispanic notation in Aem 56 (E-Mh cod 56); otherwise, (diastematic) aquitanian notation.

(x2) Chants duplicated (discarding incipits) in Silos 4. Both neume sequences are shown in the corresponding score.