I was sort of hoping that a greater scholar than I would weigh in on the matter of the Romanian letters. The relatively terse treatment of them by Cardine is disappointing. I'm sure he could have said much more. There is some discussion of them in Hiley's great Handbook..., which is the successor to Apel's work. For the most part these letters really get short shrift. Semiologists attempt to do them some justice, but as for the not-Solesmes-method folks, their existence is ignored totally.

It seems to me that the fundamental importance of these letters lies precisely in that they betray a praxis in which individual cantors and scholae were free to an extent to improvise and perform with a creative approach the sacred chanted verba. It is obvious that they were not following a rote system in which each neume had an unwavering value and method of performance but that the singers approached the chant, especially with the Romanian letters, perhaps not unlike a baroque musician would have approached the ornamentation of his day; which is to say that he was not at total unreferenced liberty to invent, but that within the parameters of 'convention' the same chant may have been performed differently by different singers. One singer may have observed tenere or mediocriter, etc., rather differently than his fellows. This chant, after all, 'belonged' to these people. They lived with and lived the language; and the language was what the chant was about. We, then, must assume that however they performed it spontaneously and off-the-cuff must surely have been quite different in sound from the more studied and codified performance methods of recent history - the moreso in that, with respect, most of these methods take no account at all of recent scholarship into paleographic evidence.

I'm sure that someone here will do this topic more justice than have I. It really deserves far greater attention.

MJO,

The letters have had me very curious as well, and I have found Hiley's Handbook and Fr. Kelly's Chant Manual both to be fairly helpful as a kickoff point. What really prompted my curiosity was noticing that some manuscripts (e.g., Einsiedeln 121) make liberal use of the letters, while in others they are rare (perhaps only the occasional celeriter). Some seem to use mediocriter either exclusively in a melodic or rhythmic context, others in both, and in some places it appears on its own where the context has to be determined. Hence my question specifically relating the position of the m with respect to the virga, because a little while later I discovered it below the virga in a different chant. So I thought that the position of the m relative to the note was likely significant.

A fascinating thing is finding in the same manuscript (e.g. St. Gall 359) some chants which are copiously marked with letters, and others which have very few. I've been wondering why this is the case.

Anyone can browse the manuscripts fairly easily from here without too much clicking.

Another source of information on the litterae significativae that should be mentioned here is Einfuhrung in die Interpretation des gregorianischen Chorals (1987) by Luigi Agustoni and Johannes Berchmans Goschl. Volume I of this three-volume work is available in an English translation by Fr. Columba Kelly. The authors are considerably more systematic and detailed in their treatment than is Cardine in Gregorian Semiology.

JonLaird,

The question of 'm' is an unsolved mystery, as far as I know. Dom Murray in the 1960s proposed that 'm' can be associated with pitch (am, im), rhythm (tm, cm), and possibly also dynamics (bm).
When it refers to pitch, e.g. 'mi' or 'im' = "not too low", as Incantu said, it tends to signify a pitch jump less than an experienced cantor is likely to make at that point, say only a step higher where one is likely to sing two steps higher.

When it's about rhythm, some say 'tm' and 'cm' can be treated as equivalent to 't' and 'c'; others say they are a nuance. I am hanging on for newer insights in the future.

As for the lone 'm'? No clear answer there too.

Let's look at Oculi mei:

O(culi). The 'm' at the top of the angular pes could be melodic ("See this 'sm' here for 'sursum mediocriter'? It means high but not too high!"), but it seems redundant when the leap of a fifth from G to d is already the highest possible leap in Gregorian chant. It could be rhythmic also ("The second note of this pes is long, not short! Don't speed into the next syllable so hastily!"). Both interpretations are possible but neither is self-evident. Given Laon and other manuscripts lack any parallel signification, I would say you can safely ignore it.

(Domi)num. The 'm' at the top of the short-short-long torculus seems to be about pitch, as M. Jackson Osborn suggests ("Remember to sing this torculus as abG, not acG!") If it is rhythmic, it probably signifies something like "This torculus is an ordinary short + short + long, not an implied long + auxiliary + long!" Dom Murray had suggested 'm' might be employed to signify "Sing it normal, not fancy." In my opinion, this is another case where both options are vaguely unnecessary.

(laque)o. The 'im' here surely signifies "not too low!" as in to drop the pes to the a, not to the G. The meaning here is self-evident.

re(spice). M. Jackson Osborn says the lone 'm' here on a virga means to rise to the d, not too high to the e. That is possible, but we cannot be sure it is what the copyist intended. Murray might say it is rhythmic, saying "ordinary long note here!", given that St. Gall seems to have such markings at the beginnings of new systems, indicating a historical tendency of the St. Gall singers to speed through the opening syllables. However, Laon's ordinary long uncinus shows that the 'm' can probably be ignored in practice.

The reason the letters are inconsistently found more here, less there, in St. Gall 359 is because they were a mnemonic aid to the singers. The easier it was to remember a melody (perhaps), the safer it was for the copyist of the manuscript to become lazy. The greater the danger of singers messing up, the more necessary the copyist deemed it to mark letters to warn them. The 'm' is one of the more inconsistent letters and is hardly found in Laon. Personally I think the letter can be ignored in practice for the most part.