CLONMELSH and MATHEMATICS Fr Ingram Memorial Lecture, Carlow, 24th November 2000 Maurice OReilly, St Patrick’s College, Drumcondra, Dublin 9 * Where is Clonmelsh? Clonmelsh is a parish about 6 or 7 km south of Carlow town, on the way to Muine Bheag. According to Thomas Fanning, Department of Archaeology, NUIG, there is little evidence, on any but the first edition of the Ordinance Survey map for County Carlow (sheet 12), of earthworks in the townland of Garryhundon identified as Killogan. Quarrying and dumping had badly damaged the site as evidenced by an aerial photograph of 1967. However, in 1892, an early granite cross was found there dating from between the late 8th and early 10th century. "The nature of the enclosure at Killogan and particularly the presence of the cross point to the existence of an early monastic site or cell, clearly one of a number in the immediate area." [1] I visited Clonmelsh on the chilly yet dry afternoon of 24th November 2000. It was indeed as Fanning had described it: a quarry with little hint of its impressive past of which more below. At the centre of Killogan burial ground which was quite overgrown, the cross which had been repaired in the mid 1980s, was again in two pieces, its moss-covered head leaning, forlorn, against the socket stone. The following morning, in a misty drizzle, I returned to the site with Elizabeth Oldham and Olivia Bree – the first IMTA fieldtrip to Clonmelsh! In this paper, I shall try to treat some of the disparate ideas about the interaction of place and history, of culture and mathematics, which have distracted me from time to time over the past four years. Clonmelsh & Rath Melsigi Today, Clonmelsh is unlikely to attract visitors in the same numbers as does, say, St Mullins or Glendalough; yet there is strong evidence that, in the 7th century, if not a popular destination, it was at least a significant centre of learning, under the name of Rath Melsigi. Indeed, Dáibhí Ó Cróinín has argued that manuscripts associated with the monastery of Echternach in Luxembourg may well have been written at Rath Melsigi [2]. Moreover, his evidence contributes to the debate, still unresolved, about the provenance of the great 'Insular' illuminated manuscripts including the Book of Durrow, the Lindisfarne Gospels and the Book of Kells. Another great manuscript of this family is the Echternach or Willibrord Gospels, now in the Bibliotèque Nationale, Paris. Even if the Echternach manuscripts were not actually written in Carlow, Ó Cróinín argues that Rath Melsigi was their "most natural historical setting". How could this have been? References in Bede Bede, the Venerable, lived from about 673 to 735 apparently without leaving Northumbria. In book III, chapter 27 of his Ecclesiastical History of the English People [3], he wrote: * e-mail: maurice.oreilly@spd.ie In this year of our Lord 664 there was an eclipse of the sun on 3 May about 4 o'clock in the afternoon. In the same year a sudden pestilence first depopulated the southern parts of Britain and afterwards attacked the kingdom of Northumbria, raging far and wide with cruel devastation and laying low a vast number of people. Bishop Tuda was carried off by it and honourably buried in the monastery called Pægnalæch. The plague did equal destruction in Ireland. At this time there were many in England, both nobles and commons, who, in the days of Bishops Finan and Colman, had left their own country and retired to Ireland either for the sake of religious studies or to live a more ascetic life. In course of time some of these devoted themselves faithfully to the monastic life, while others preferred to travel round to the cells of various teachers and apply themselves to study. The Irish welcomed them all gladly, gave them their daily food, and also provided them with books to read and with instruction, without asking any payment. Among these were two young Englishmen of great ability, named Æthelhun and Egbert, both of noble birth. The former was a brother of Æthelwine, a man equally beloved of God, who, later on, also went to Ireland to study; when he had been well grounded he returned to his native land and was made bishop in the kingdom of Lindsey, over which he ruled for a long time with great distinction. Æthelhun and Egbert were in a monastery which the Irish call Rath Melsigi, and all of their companions were carried off by the plague or scattered about in various places, while they themselves were all stricken by the same disease and were dangerously ill. The account continues to explain that Æthelhun perished but Egbert was spared. Egbert intended to follow the footsteps of Wilfred (of York) in the mission to the Frisians who lived in the flat marshy land between the Rhine and the North Sea [4]. He was unsuccessful in this mission, as was his successor Uuichtberct. After Uuichtberct, Willibrord left for Fresia in 690, founded Echternach (in 699) and, under the patronage of Pippin II, lead the conversion of the Frisians, the Danes, the Thuringians and more besides [5]. The key point Ó Cróinín makes is that although the Frisian mission began from Northumbria, Egbert, Uuichtberct and Willibrord set out on their mission from Rath Melsigi. Learning in early medieval Ireland In the Late Roman Empire, the imperial curriculum comprised seven disciplines: grammar, rhetoric, logic, arithmetic, geometry, astronomy and music. The first three of these known as the trivium, the last four, the quadrivium. During this period, many textbooks were compiled such as Martianus Capella's The Marriage of Philology and Mercury (410-429), and the works of Donatus and Priscian [6]. In 410, the Visigothic leader, Alaric, sacked Rome. Ireland, having been outside the Empire, was about to undergo momentous, yet peaceful, change. We tend to think of learning in Ireland as beginning with the coming of Christianity and Latin probably in the late 4th century. This is, of course, a simplification: these two foreign elements arrived in a society enjoying a rich and structured cultural framework which was propagated orally since writing had not yet been introduced. This civilisation, by all accounts, embraced the new written learning, secular and, especially, religious from the ailing Roman Empire. After about two centuries of taking root (from, say, 400 to 600), a new scholarship flourished in Ireland throughout the next two centuries and gradually waned in the 9th century [7]. Evidence of the Irish 'Golden Age' is plain to see in metalwork, stone carving and illuminated manuscripts [8]. But there is more besides: the enormous legacy of this period found in manuscripts scattered throughout Europe bears testimony to Irish zeal for travel or peregrinatio, and learning or ubera sapientiae [9]. The Irish curriculum in this Golden Age was dominated by three areas of study: grammar, which concerned all aspects of the study of Latin, exegesis, which included biblical, liturgical and patristic studies as well as hagiography, and, finally the computus, the study of the calendar. In all of this there are two areas especially where it is fruitful to look for mathematics: the geometry of the illuminated manuscripts and the computus. What geometry did the illuminators know? Let us return to Clonmelsh, or at least to one of its leading alumni, Willibrord, who studied there for twelve years from 678 to 690. Ó Cróinín [2] suggests that Rath Melsigi is at least as good a candidate as anywhere in Northumbria for the scriptorium where the Echternach Gospels were written. He offers various historical evidence to support his claim, rather than the arguments from archaeology and art-history which had been used to suggest a Northumbrian provenance. He quotes Lowe that the manuscript belonged to Saint Willibrord himself. As mathematicians, our interest is in how the illuminators conceived geometry. Whether or not they knew, for example, Pythagoras' theorem we may never know; that they had at least a practical flair for geometry is certain. Several authors have investigated the geometrical forms in the Insular manuscripts, but none so comprehensively as Robert D. Stevick, Professor of English at the University of Washington, Seattle [10]. According to Stevick, his interest arose 'as an attempt to understand the Anglo-Saxons' divisions of their longer vernacular poems'. He continues, 'this study evolved into a dual investigation of form in art and poetry of the same locale and time'. He emphasises the possibility of practical reconstructions of geometrical forms which underlie, for example, illuminated pages from Insular manuscripts, while acknowledging that any such proposed form was not necessarily used by the artist. Unfortunately, evidence of compass marks in the vellum is unusual, an interesting exception being folio 26v. of the Lindisfarne Gospels. He also points out that individual artists were not necessarily familiar with the use of mathematics in any sophisticated way, nor with mathematical proof: 'Geometers' techniques were being used for creating forms, not for mathematical inquiry or problem solving'. His project is 'to recover the repertory of construction techniques that were used to create the forms'. I have used Maple to examine Stevick's proposal for the Imago Leonis page (75v.) of the Echternach Gospels. Here is the final result: The page itself, as its title suggests, has a wonderful rampant lion positioned diagonally across it. On this page, Stevick identifies occurrences of ratios involving 2 and the golden ratio, . Indeed, he claims that the entire geometrical form of the page can be constructed using these two 'true measures'. Without more attention to the details of draughtsmanship, it is difficult to take into account the thickness of the plan outline, which, on the original, is probably about 2.5mm. Possible discrepancies arising from ignoring this thickness can give rise, in the most extreme cases, to errors in estimating linear measurements of about 15%. With such a possible margin of error, it is difficult to distinguish between the measurements 2+2 = 3.414… and 1+22 = 3.828…, for example. A computer algebra package with good graphics such as Maple or, The Geometer's Sketchpad makes investigation of the detail of such analysis more controllable, and may well lead to new insights into geometry in Early Medieval Ireland. Learning in Europe in the 'Dark Ages' Mathematics is considered by many as a cumulative discipline, one in which theories are built up in painstakingly upon solid foundations involving axioms and theorems bound together by the super glue of logic. Fashions may come and go - conic sections will probably never again enjoy the popularity of the days of George Salmon's textbooks [11]. Some ideas such as the differential calculus are here to stay in spite of having severe doubts cast on their philosophical basis by George Berkeley 70 years after its discovery [12]. Euclid laid solid foundations for geometry about 300BC, and, although the parallel postulate was questioned repeatedly leading to non-Euclidean geometries in the 19th century, the rigour of the Greeks is still used as an exemplar. The Greeks knew how to state axioms and formulate theorems - all who engage in the true art of mathematics should do likewise! It is interesting that the real line was not axiomatized until 1872 independently by Dedekind and Cantor. In his wonderful book, Journey through Genius [13], William Dunham discusses twelve great theorems from the 5th century BC to the late 19th century. In this celebration of the great moments of mathematics, there is a gap of almost 1500 years between Heron's formula for triangular area and Cardano's solution of the cubic. This glaring time-gap is known in the history of mathematics and elsewhere as the Dark Ages when classical knowledge was lost to the Western world. Yet we know that mathematics is more than the rigour of axioms and theorems, however important they may be. Beyond the south-eastern extremities of Europe, in Baghdad, algebra arose in the 9th century [14]. It was to have a major impact on the future development of European mathematics in spite of Europe's obsessions with its Hellenic legitimacy. (See Cifoletti, Høyrup and others in [15].) I believe that the mathematics practised in the north-western extremities of Europe, in Ireland in the 7th and 8th centuries deserves more attention, if not among the global mathematical community, at least among those interested in mathematics in Ireland and its neighbours. The calendar & the computus Now let us turn to the question of the calendar. Today we take the calendar for granted, accepting details like months of differing lengths and leap years without a second thought. Because of millennium fever, quite a few popular books have been published recently on the calendar. EG Richards’ book, Mapping Time, contains a good deal of interesting material including several algorithms for calendar calculations [16]. The final part of this book is on Easter, the issue which, above all, motivated the study of the computus in the early Christian era. The two units of time which are the main concern of the computus are the solar year and the synodical or lunar month. Nowadays, the solar month is defined as the period of time during which the earth makes one revolution around the sun, measured between two successive vernal equinoxes, equal to 365.242199 days. On the other hand, the synodical month is the period of time taken by the moon to make one complete revolution around the earth, measured between two successive new moons, equal to 29.53059 days [17]. Thus the solar year is 12 synodical months and 10.875 days The Easter question arose from the Jewish calendar in which the date of the Passover (or Pesach) depended on the lunar rather than the solar calendar. In the early church, the question of universal agreement on the date of celebration of Easter caused much controversy. There were theological issues such as deciding on which day of the week Easter should be celebrated, or, having decided on Sunday, what the permitted range of dates for Easter should be. These were not the only questions to be resolved; there was also the technical puzzle of how to predict the relationship between the solar and lunar calendars in advance [18]. Returning to Clonmelsh, or Rath Melsigi, Ó Cróinín [2] argues a strong case that a particular Echternach manuscript (again in the Bibliotèque Nationale, Paris) containing much computistical material, was written there. We can be confident that the computus was a significant element in the education of students at Rath Melsigi. The technical computistical challenge involved the establishment of a periodic cycle to model the solar and lunar calendars. The relative position of the sun and moon would be approximately the same at the end of each cycle as at its beginning. Moreover, a cycle, just like the familiar sine function’s cycle of period 2 radians, represented the characteristic pattern of the intertwined motions of sun and moon. By the middle of the 7th century there were two main contenders: a 19-year cycle favoured by both Alexandria and Rome and an 84-year cycle which still enjoyed support in parts of the British Isles. In the 19-year cycle there were 19365 = 6935 days (excluding the bissextiles or leap days, one every four years). These 19 years comprised 12 common years, each of 12 lunar months, together with 7 embolistic years, each of 13 months. In each year, there were 6 ‘hollow’ months of 29 days together with 6 ‘full’ months of 30 days, while the extra embolistic month was also full. Adding all this up yields 12354 + 7384 = 6936 days, or one day too many. This error was rectified by the saltus lunae which shortened one of the full months by one day, thus making it hollow. In a leap year, the month containing the 24th February (or the 6th kalends of March) was given an extra day, the ‘second sixth’ or bissextus in both the solar and lunar calendars without causing any disruption to the cycle. Likewise, the 84-year cycle of 84365=30660 days comprised 53 common years and 31 embolistic years, with 6 salti lunae to correct the discrepancy. The Synod of Whitby of 664 – the same year as the pestilence – brought victory to the camp of the 19-year cycle. The so-called Padua latercus (or Paschal table) which was rediscovered in 1985 by Ó Cróinín [19], having been lost for about 1300 years, brought to light the details of the 84-year cycle. Following this discovery and noting its complete dependency on a text called De ratione paschali, attributed to the third century Anatolius of Laodicea, McCarthy [20] has suggested that the corruption of tables in the latter may have been deliberate and politically motivated. Conclusions It is a difficult task for those enjoying riding the Celtic Tiger, or indeed those hanging on by his tail, to begin to appreciate what mathematical activity might have been conducted by the Irish and their circle over 1300 years ago. For one thing the notation was different: it was to take over three more centuries before Arabic numerals were introduced to Western Europe. Nonetheless, I believe that the material available to us contains evidence of mathematics at least of an algorithmic nature. By taking an interest in this evidence, we gain important insights to pass on to our students, not least a feeling for how mathematics and culture were inter-related in our past, and continue to be today. The spirit of the Irish ‘Golden Age’ offers much to inspire us: an appreciation of learning, including the geometric intricacies of the illuminated manuscripts, the arithmetic – almost algebraic – complications of the computus, and generous hospitality at Clonmelsh which fostered it all. References 1. Fanning, T, Some Field Monuments in the Townlands of Clonmelsh and Garryhundon, Co Carlow, Peritia 3 (1984) 43-49. 2. Ó Cróinín, D, Rath Melsigi, Willibrord, and the Earliest Echternach Manuscripts, Peritia 3 (1984) 17-42. 3. Bede, The Ecclesiastical History of the English People, J McClure & R Collins (eds), Oxford UP, 1999. 4. Collins, R, Early Medieval Europe 300-1000, Macmillan, 1991. 5. Fletcher, R, The Conversion of Europe from Paganism to Christianity 371-1386, Fontana Press, 1998. 6. Fossier, R (ed), The Cambridge Illustrated History of the Middle Ages 350-950, CUP, 1997. 7. Ó Cróinín, D, Early Medieval Ireland 400-1200, Longman, 1997. 8. Harbison, P, The Golden Age of Irish Art, Thames & Hudson 1999. 9. Richter, M, Ireland and her Neighbours in the Seventh Century, Four Courts Press, 1999. 10. Stevick, RD, The Earliest Irish and English Bookarts - Visual and Poetic Forms before AD 1000, Univ. Penn. Press, 1994. 11. Houston, K (ed), Creators of Mathematics - The Irish Connection, UCD Press, 2000. 12. Fauvel, J & Gray J, The History of Mathematics - A Reader, Macmillan, 1987. 13. Dunham, W, Journey through Genius - The Great Theorems of Mathematics, Wiley, 1990. 14. Rashid, R (ed), Histoire des sciences arabes, 2 - Mathématiques et physique, Seuil, Paris 1997. 15. Goldstein, C, Gray & Ritter (eds), L’Europe mathématique/ Mathematical Europe, Editions de la Maison des sciences de l’homme, Paris 1996. 16. Richards, EG, Mapping Time – The Calendar and its History, OUP 1998. 17. Encyclopædia Britannica, 15th edn (1978), Micropædia, vol. 3, pp595-599. 18. McCarthy, DP, Easter principles and a fifth century lunar cycle used in the British Isles, J for the History of Astronomy, 24 (1993), pp 204-24. 19. McCarthy, D & Ó Cróinín, D, The Lost Irish 84-year Easter Table Rediscovered, Peritia, 6-7 (1987-8), pp227-42. 20. McCarthy, DP, The Lunar and Paschal Tables of De ratione paschali Attributed to Anatolius of Laodicea, Archive for History of Exact Sciences, 49 (1996), pp 285-320. Acknowledgement The author expresses his gratitude to Dr Dan McCarthy, Trinity College, Dublin, for drawing attention to certain misleading passages in an earlier draft of this paper. Any such passages that remain are, of course, entirely the responsibility of the author!