Music For The Divine Office
On this disc we meet Tallis as a composer of choral music for the Divine Office, the cycle of eight services - Matins, Lauds, Prime, Terce, Sext, None, Vespers, and Compline, known collectively as the Canonical Hours - sung daily by communities of religious in Latin Christendom.
Nothing about Tallis's early career suggests that he was destined to reach the top of his profession. Nevertheless, scarcely more than ten years separate his first known musical appointment, which was extremely humble, from his last, which could not have been more prestigious. In 1532 he was organist of the small Benedictine priory at Dover - a minor post if ever there was one. Five years later he had moved to London, where he was employed either as a singer or as organist by the parish church of St.Mary-at-Hill, which was noted for its music. in 1538 he abandoned London for the apparent security of a permanent appointment as a member of the Lady Chapel choir of the Augustinian abbey of Holy Cross at Waltham in Essex. However, the dissolution of the abbey in March 1540 left him once again without employment.
In the spring of 1540 the prospects for an unemployed church musician cannot have seemed promising, but Tallis now succeeded in making what was possibly the most consequential move of his career. A fortnight after Waltham Abbey was dissolved, Canterbury Cathedral ceased to be a Benedictine monastery; it was reorganised with a secular dean and chapter and provided with an enlarged choir consisting of ten boys and twelve men, worthy of the cathedrals status as the fons et origo of a national church. Tallis joined the new choir during the summer of 1540 and remained one of its senior members for two or three years. These years must have been lively, not only because of the challenge of quickly assembling an impressive and extensive repertory, but also on account of the fierce disputes that arose between the conservative cathedral dignitaries and their reform-minded archbishop, Thomas Cranmer.
It could have been through Cranmer, Henry VIII's most trusted counsellor during his declining years, that Tallis gained a place in the royal household chapel; the exact date of his appointment is not known, but his name occurs about half-way down the list of Gentlemen (or singers) of the Chapel in the Lay subsidy roll of 1543/4. He remained a Gentleman of the Chapel for the rest of his life, rising in seniority until he became its acknowledged doyen. He may have acted as chapel organist throughout this time, although he is not referred to in this capacity until the 1570s. Apart from playing the organ, Tallis's main duty during his early years in the royal household chapel was probably the composition of music. This would have been an important responsibility, because the chapel was undoubtedly expected to demonstrate how government policy on religion should be interpreted in practice. When, during the 1540s, Tallis and his colleagues embarked upon the composition of polyphony for the Divine Office, it must surely have been in response to an official policy of counteracting reformist criticism of church music by integrating it better into the liturgy.
The Divine Office originated during the early days of monasticism in order to provide a format for communal psalm-singing, audition of Scripture, and prayer. Like the Mass, the Office services became more elaborate as they evolved: the chanted psalms were preceded and followed by plainchant antiphons; the readings were followed by plainchant responsories; each service included a plainchant hymn; and the more important services also included an Old or New Testament canticle. Despite these developments, the services of the Office never developed an intricate ceremonial to rival that of Mass; those that came nearest were Matins and Vespers, and to a lesser extent Lauds. Although it remained the type of monastic worship par excellence, reaching its highest pitch of development among the Cluniacs and Benedictines, the Divine Office was adopted by the medieval secular church wherever-as in cathedral and other collegiate foundations a resident body of clergy was available to perform it. During the later Middle Ages it penetrated the secular world still further. An abbreviated version of the Office consisting of a reduced cycle of services was sung in the household chapels of the aristocracy; under the Lancastrian and early Tudor kings, for example, the royal household chapel regularly sang Matins, Lauds, Prime, Vespers and Compline, but apparently did not observe the other Hours. Selected services from the Divine Office also formed the essential contents of the book that became part of the standard equipment of private lay devotion - the Book of Hours.
If the evidence of the surviving musical sources is to be trusted, early Tudor composers were like their predecessors in producing relatively little polyphony for the Divine Office. Instead they concentrated upon music for the Mass, in the form of the cyclic Mass, and for the post-Compline devotion, in the form of the votive antiphon. Mass was the service with the most powerful spiritual charge, the highest public profile and the largest musical content; the post-Compline devotion was a paraliturgical invocation of Our Lady, Our Lord or a favourite saint in whose intercessory potential institutional and individual patrons alike made lavish investment. Tallis's own contribution to these two genres, for example the votive antiphon Salve intemerata and the cyclic Mass based upon it, has been explored in Volume 1 in this series. The only Office item to which English composers had traditionally paid so much attention was the Magnificat, the New Testament canticle sung at Vespers, the earliest polyphonic settings of which date from the later fourteenth century. It is probably significant that, of all the Office services, Vespers had the most ornate ceremonial and the strongest tradition of attendance by the laity, both high- and low-born. Even so, the number of surviving Magnificat settings represents a very small proportion - at a guess, less than a tenth of the total corpus of early Tudor Church music. Of other polyphony for the Office, there is hardly a trace.
Two compositions of the Magnificat by Tallis survive: the one in five voices was included in Volume 2 of this series, and the four-part setting is recorded here. There are several reasons for believing this four-part Magnificat to be a very early work. First, its clumsiness (particularly its angular vocal lines, random dissonance and crowded textures) implies inexperience. Secondly, it conforms almost completely with the standard scheme for setting the Magnificat that English composers had followed since the mid-fifteenth century but were beginning to abandon in the 1530s. Only the six even-numbered verses are set, the first two in triple metre, the next two in duple and the last two in triple again. The first, third and fifth verses are fully scored, while the second and the first halves of the fourth and the sixth are for reduced forces. The setting is based upon a rather unusual kind of cantus firmus called a faburden, not itself a plainchant, but a melody that had originally been devised as a counterpoint to a plainchant as part of the process of improvising polyphony. For some reason, the faburdens to the Magnificat tones (the plainchant formulae to which Magnificats were usually chanted), rather than the tones themselves, came to be used as the cantus firmi of composed Magnificats. Tallis's cantus firmus is the faburden to the first Magnificat tone: its opening notes, descending stepwise through a minor third, can be heard clearly at the start of every polyphonic verse; the other verses are sung to the tone itself.
Towards the end of Henry VIII's reign, composers began to produce choral settings of Office responsories and hymns. Precise dates are difficult to establish, but the senior composers associated with these developments seem to have been John Taverner (d.1545), whose musical employment (but not necessarily his activity as a composer) evidently came to an end in the later 1530s, and John Redford, who died in 1547. The new fashion was continued mainly by two younger composers:Thomas Tallis, by whom we have seven hymn settings and nine responsories, and John Sheppard, with about seventeen hymns and twenty responsories. The impulse behind these innovations, whether emanating from the composers themselves or - much more likely - from their employers, must surely have been religious, reflecting the hostility towards the votive antiphon (with its implicit acceptance of the existence of purgatory and the efficacy of indulgences and intercession) current in some circles within the Anglican Church. Most responsories took their texts from the psalter or from the Gospels and thus had unimpeachable scriptural credentials, while the antiquity and irrefutable orthodoxy of many of the Office hymns could have made them seem the next best thing to scripture.
It seems most likely that Tallis composed his hymns and most of his responsories after joining the royal household chapel. In view of the religious climate at court during the final years of Henry VIII, when despite his own religious conservatism the ageing king allowed the reformers rather more scope than he had permitted a few years earlier, it seems reasonable to interpret the introduction of these genres as an attempt at compromise: it enabled elaborate polyphonic composition to continue within the Latin liturgy while disarming the kind of criticism to which the votive antiphon was vulnerable. After the Protestant interlude under Edward VI, the composition of hymns and responsories was resumed, perhaps on a lesser scale, under Mary; there are a few examples by composers such as Robert Whyte and William Mundy, who were born too late to have been active during her father's reign. Although any attempt to assign Tallis's works in these forms either to Henry's reign or to Mary's must be conjectural, it is not impossible that the majority of them are late Henrician. The small scale solo responsories Hodie nobis caelorum, In pace in idipsum and Audivi vocem, however, must date from earlier in Tallis's career. These are three of the very few responsonsorial items that earlier Tudor composers had occasionally set in polyphony, probably because each of them was in some way or other liturgically or ritually unusual. Tallis's approach and style in these works suggest that he was continuing a tradition rather than innovating.
From a liturgical point of view, Tallis's hymns and responsories form a rather coherent group. Four hymns (Salvator mundi, Jesu salvator sacculi and the festal and ferial settings of Te lucis ante terminum) and one responsory (In pace in idipsum) are for Compline; taken together, they provide polyphony for Compline during most of the year, and were perhaps intended to compensate for the disappearance of the polyphonic votive antiphon occasioned by the abandonment of the post-Compline devotion. Two other hymns (Quod chorus vatum and Jam Christus astra ascenderat) and three responsories (Videte miraculum, Loquebantur variis linguis and Homo quidam fecit) are for major feast days (Purification, Pentecost and Corpus Christi) at first Vespers, the service with which the celebration of a major feast day began. The remaining four responsories (Hodie nobis caeloum, Dum transisset sabbatum, Honor virtus et potestas and Audivi vocem) are for Matins on the major feasts of Christmas, Easter, Trinity and All Saints respectively. It is noteworthy that all of the feast days mentioned - Christmas, Purification, Easter, Corpus Christi, Pentecost, Trinity and All Saints - were occasions on which the Tudor monarchs traditionally took a prominent role in the royal household's religious observance. The remaining hymn (Sermone blando, for Lauds from Low Sunday until Ascension) and responsory (Candidi facti sunt, for one or more apostles, or an evangelist, in Eastertide) are harder to account for . Sermone blando was also used by Ludford as the cantus firmus of a Mass now lost; possibly the chant had a significance now unappreciated. The only apostles or evangelists whose feasts can fall within Eastertide are St.Mark (25 April) and Sts Philip and James (1 May), so perhaps Candidi facti sunt was written for a year when one or other of these relatively minor feasts coincided with an important court occasion.
Tallis's approach to composing hymns and responsories is methodical but inventive. The hymns are essentially settings of the original plainchant melodies. When sung entirely in plainchant, hymns were performed alternatim, the two sides of the choir singing alternate verses to the same melody. Tallis preserves this alternatim structure by setting only the even-numbered verses and usually also the doxology in polyphony, leaving the other verses to be sung to the original chant. Quod chorus vatum thus has two polyphonic verses, while Salvator mundi, Jesu salvator saeculi and Jam Christus astra ascenderat have three and Sermone blando has four. Quod chorus and Salvator mundi have different music for each polyphonic verse; Jesu salvator and Jam Christus repeat the music of the first verse for the second ( with some small changes in the case of Jam Christus ) and have new music for the third; and in Sermone blando the music for the first verse is repeated for the second, and that for the third is repeated for the fourth, with two of the inner voices exchanging parts each time. Tallis always sets the first polyphonic verse or pair of verses in compound duple metre and the others in simple duple. The choice of the former ( six-eight time, in other words ) is surprising for, although this metre had been common enough in English music during the early fifteenth century, it had subsequently fallen out of fashion and had been in disuse for more than a hundred years. Tallis may perhaps have chosen it in order to reproduce the effect of an oral tradition of singing plainchant hymns metrically.
It is easy to underestimate the craftsmanship and ingenuity of these hymn settings. They are all in five voices, with the plainchant in the top voice, but the variety that Tallis can achieve despite what might be considered a mechanical approach is quite astonishing. The simplest of the settings is Quod chorus vatum, in which the plainchant melody is sung virtually unadorned over a loosely imitative four-part texture. In Sermone blando the highest part again sings the original melody, this time without any decoration at all, but each line of the melody is also alluded to by an increasing number of the lower voices: in the first polyphonic verse, for example, the first line ( 'Illae dum pergunt concitae' ) is anticipated by the first contratenor, the second ('Apostolis hoc dicere') by the tenor, and the first contratenor, the third ( 'Videntes eum vivere') by both contratenors and the tenor, and the fourth ('Osculantur pedes domini') by all four lower voices. In the other settings recorded here Tallis creates a cumulative effect by a variety of means. In Jesu salvator saeculi the voices begin in complete rhythmic unanimity, but thereafter the lower voices become increasingly independent rhythmically, the vocal scoring evolves from simple block contrast to constant subtle changes of colour, and the setting ends with an extended and beautifully balanced 'Amen'. Jam Christus astra ascenderat demonstrates another kind of unobtrusive craftsmanship: in the first two verses Tallis works the plainchant in canon in the treble and contratenor, while in the third the chant sails over an independent imitative texture that becomes ever more tightly argued as it proceeds. Salvator mundi is perhaps the most impressive of these settings in terms of planning: the highest voice decorates plainchant with increasing profusion in successive verses, and in each verse the lower voices weave an imitative accompaniment based on an apparently inexhaustible fund of new ideas.
Like his hymns, Tallis's responsories reflect their plainchant origins, in that their performance involves the alternation of plainchant and polyphony, and that the polyphonic sections habitually quote the plainchant that they replace. A plainchant responsory is a lengthy item following a reading, allowing time for the edifying words to be contemplated. It usually consists of the responsory itself, begun by soloists and continued by the choir, a verse sung either by the beginners or by another group of soloists, and the Gloria patri, sung by the singers of the verse; the choir repeats the last two sections of the responsory after the verse, and the last section again after the Gloria patri. Tallis and his contemporaries employed two contrasting methods of setting responsories: either they set the solo portions in polyphony and left the choral sections in plainchant, or they set the choral sections and left the solo portions in plainchant. The first method produces what is known as the solo responsory, while the second results in the choral responsory. The solo responsory was the traditional type, traceable at least as early as Leonin's responsorial compositions for Notre Dame de Paris in the mid-twelfth century. The choral responsory may well have been the invention of Taverner, appearing for the first time in his settings of Dum transisset sabbatum. Among Tallis's responsorial works, Hodie nobis, In pace in idipsum and Audivi vocem are solo responsories, and Videte miraculum, Dum transisset sabbatum, Loquebantur varriis linguis, Homo quidam fecit, Candidi facti sunt and Honor virtus et potestas are choral responsories. The two groups differ also in other ways: the solo responsories are for four voices, are on a fairly small scale and scatter references to the chant throughout the polyphonic texture, whereas the choral responsories are for five, six or seven voices, are on a monumental scale and build the plainchant into the polyphony as a monorhythmic cantus firmus quoted continuously by a single voice.
Tallis's responsories also resemble his hymns in showing outstanding imagination in the detailed application of a basically standard approach. Of the two solo responsories recorded here, Hodie nobis caelorum has very modest dimensions and makes only the most fleeting of references to the chant, whereas the thorough imitative discussion of the more extended In pace in idipsum is based on a series of melodic ideas derived from the chant. It is interesting that Tallis fails to make provision for the traditional way of singing for the verse 'Gloria in excelsis' of Hodie nobis caelorum, which in the Salisbury rite was was sung by five boys , clothed in white and holding lighted tapers, from a high place beyond the high altar. Tallis's setting, however, is in four parts, not five, and the written pitch implies performance by men. On this recording the piece is sung twice, once by high and once by low voices. In the choral responsories the plainchant forms the backbone of the polyphonic texture, being stated without decoration in equal note values somewhat larger than those of the other voices; in the five-part Dum transisset sabbatum it is at the top of the texture, while in the six-part Videte miraculum
and the seven-part Loquebantur variis linguis it is in the middle. The other voices in the texture spin their own cotrapuntal web, sometimes (for example, in the opening bars of Dum transisset and Loquebantur, and at 'miraculum', 'matris' and 'quae se nescit' in Videte) making reference to the chant melody, and at other times discussing their own independent material. If Loquebantur impresses immediately on account of its seven-part texture (surely symbolising the sevenfold gifts of the Pentecostal spirit) and its virile melodic material, the hypnotic Videte reveals its secrets rather more slowly. Conceived on an enormous scale, Videte evinces remarkable imagination and a masterly sense of timing; by means of continual deft adjustment of pace, melodic outline, vocal colour and level of dissonance, Tallis leads us into the same world of timeless truth that we glimpse in Spem in alium.
Nick Sandon, 19 August 1998