Reconstructing Roman Marching Tunes

Crispvs (Paul Geddes)

One thing which distinguishes the RMRS from many other groups is that we have a good repertoire of songs and chants which we can use either whilst marching or during our field displays. This adds to the effectiveness of our impression and is something which we are often complimented on by members of other groups.

One problem which our enthusiastic singing and chanting perhaps hides though, is the fact that we have far less evidence for the songs of the period we depict than exists for other. later, periods. What evidence there is amounts to little more than a thimble full compared to the bucket loads available to, for instance, Napoleonic period re-enactors. In our repertoire we have just a mere four songs and two chants. Admittedly this is enough for the needs of our displays, but it is not much in the grand scheme of things. To make matters worse, fully half of our repertoire is not genuine Roman material but is instead recently written material which has been translated into Latin for us, precisely because the surviving sample is so small. That is not to say that we should not be thankful for the work of those who wrote and translated the modern reconstructions. On the contrary, we have cause to be very grateful, as their work makes us much richer as a group and adds immeasurably to our display.

What we use from Roman times however, is limited to one chant ('Cras amet' - actually the refrain from a long poem 'Pervigilium Veneris', possibly by Tiberianus) and two songs, namely 'Bacche' (which is a poem by Florus), and 'Urbane', the ribald song (or perhaps songs) quoted by Suetonius as having been sung during Julius Caesar's triumph. Of these, only 'Urbane' is known to have been sung by marching soldiers.

Another area we are lacking in is genuine Roman tunes to sing our songs to. Obviously, the Romans had tunes, but they lacked the modern technology we have to record sound and although the ancient world did have systems of musical notation they left us no records of any marching tunes. Therefore we have had to make up or find tunes of our own to sing our Latin songs to. For this, again, we owe a debt of gratitude to those who created or found the tunes we use.

Do we in fact know anything about genuine Roman marching tunes? Perhaps surprisingly, the answer is: yes - we do. Although we do not have any actual tunes, we do have a number of clues to how Roman marching songs might have sounded.

The first of these is a mention by the poet Prudentius in his 'Psychomachia', where he states that Roman soldiers marched to a meter of three groups of four syllables followed by a group of three syllables (total: fifteen syllables per line). I checked this rhythm against the song (or songs) quoted by Suetonius and found that four of the seven lines do indeed follow this syllabic pattern. Add to that the possibility that the lines which now have sixteen or seventeen syllables may originally have featured some poor grammar which has been corrected by the well educated Suetonius (thus changing the syllabic structure) and we have the possibility that all seven lines may originally have had fifteen syllables. Interestingly, 'Bacche', the other song in our repertoire dating from ancient times, also features this syllabic structure and therefore the rhythm we are familiar with for that song would also be applicable to 'Urbani'. Count: ONE, two, three, four - FIVE, six, seven, eight - NINE, ten, eleven, twelve - ONE, two, three; and you have it.

Next we have the fact that although no music is known to survive from the first century AD period we normally depict, there IS music which has survived from what we think of as the late Roman period. To modern ears the plainsong chants used by the Catholic church and Church of England have an unmistakably ecclesiastical sound, often described as 'Gregorian chant' (although 'Gregorian' chant is only one of a number of types of plainsong). However, the earliest of these plainsong chants actually date to the fourth century AD and thus plainsong actually preserves an authentic Roman style of singing. It could be argued that plainsong was a purely religious style but the evidence for that is lacking, as no other style survives to compare with it. The early church may have sung in what was a perfectly ordinary style of singing for the time. Liturgical clothing was originally little different to normal formal clothing after all. Some of the types of ecclesiastical vestments which have lasted in use right up to the present day, such as the alb, were simply the normal formal clothing which had largely superseded the toga by the fifth century AD. Early Christian altars also appear to have been very similar in shape to traditional Roman ones, the only obvious difference being that they were covered by a white cloth. If the early church was happy to dress its priests in the equivalent of a modern business suit and use altars of a normal Roman form, is there any reason to suppose it would have used anything other than a normal contemporary style of music for its songs? Probably not. Therefore it is reasonable to postulate that normal Roman singing of the time may often have taken the form of a rhythmic near-monotone tune with each line finishing in a tone which rose, fell and rose again, which repeated identically with every line including the final one, as is still familiar to many people who attend Catholic or Church of England churches today. Next we might rhetorically ask the unanswerable question of whether this style of music was new at the time, or already had a long history with the Romans.

Our next piece of possible evidence extends somewhat from the last. Of the modern languages which have developed from Latin, Italian and Spanish remain the most pure. Italian is today spoken with a great deal of tonal variation but in contrast to this, Spanish is often spoken with very little tonal variation until the end of the sentence, where the tone typically rises and then falls again. In itself this may mean nothing, but it is interesting to note that the modern tonal arrangement of Spanish can appear similar to that of plainsong, which also often exhibits little or no tonal variation until the end of a sentence.

The final piece of possible evidence to present here for the sound of Roman singing is the fact that our modern word 'chant' comes to us directly from the Latin word 'cantere' ('to sing' - also preserved in the French word 'chanter', again - 'to sing'). Was Roman singing typically in the form of rhythmic chanting, with tonal variation into what we would think of as 'tunes' being a later introduction from elsewhere? Again, we cannot satisfactorily answer this question but it is worth considering nonetheless. Although we know the Romans must have been used to tunes played on flutes, reed instruments and harps, this does not necessarily mean that they must have sung to such tunes. After all, if they were careful to reserve particular artistic styles for specific materials and classes of object, as we know they did, might they not have been comfortable with similar divisions in their music? What ancient poetry (and probably, by inference, song) is rigidly concerned with is meter. Might the use of particular meters largely have negated the need for sung 'tunes' for the traditionally minded Romans?

We return now to our one surviving genuine Roman marching song - 'Urbani'. For years now we have sung it to an old Latvian army tune which Juris which, in addition to sounding quite good, does certainly fit the words reasonably well. However, this tune is of course a modern one and on a number of occasions I have noticed new members of the unit and members of the public commenting that the tune reminded them of a First World War marching song, which is what it is of course. Without trying to take anything away from a highly respected member's contribution of a tune which has worked well for us for around two decades, hopefully I have shown through this short article that we actually do have sufficient evidence to try to arrive now at what may be a more authentic Roman style of tune for this song. There is, of course, the caveat that we cannot guarantee that the music of the fourth century AD was like that of the first century AD, but there is always the possibility that it may have been. Whatever the case, the fact remains that the music of the fourth century AD is fifteen hundred years closer to what the real sound may have been like than the tune, good as it may be, that we currently use.


Schwartz, C (Ed.), Chambers Dictionary, Chambers Harrap Publishers Ltd, Edinburgh, 1993

Duff, JR and AM (Trans.), Minor Latin Poets Vol. II, Loeb Classical Library 434, Harvard University Press, 1934

Handford, SA & Herberg, M, Pocket Latin Dictionary, Langenscheidt, Berlin, 1966

Kelly, TF, Plainsong in the Age of Polyphony, Cambridge University Press, 1992

Pocknee, CE, Liturgical Vesture - Its Origins and Development, AR Mowbray & Co, London, 1961

Raby, FJE, A History of Christian Latin Poetry from the Beginnings to the Close of the Middle Ages, 2nd Ed, Oxford, 1953

Suetonius, Divus Iulius, Suetonius Vol I, Loeb Classical Library 31, Harvard University Press, 1914

Wilson, NS, Spanish, Teach Yourself Books, 1976