A simply spectacular performance which earned the recording a Gramophone 'Editor's Choice' plaudit.
His early career was at Dover Priory (around 1532, when he would have been in his late twenties), St Mary-at-Hill, London (1537-8), and Waltham Abbey, Essex (1538). Upon its dissolution he moved in about 1541 to the reformed Canterbury Cathedral, whence he moved to be a Gentleman of the Chapel Royal in 1543; these were uncertain times indeed for musicians. He was therefore already at the Chapel by the time of Edward VI (1547), a time when there was an urgent need for music to English texts for the Edwardine reformed service. Like his contemporaries, Tallis stayed on during Maryʼs reign, when there was a renewed need for Latin music (not least to replace material which had been thrown away or wantonly destroyed during Edwardʼs reign) and then again into Elizabethʼs reign (1558), when the reformed rite was established again (from 1559). Although there were some examples of Catholic Gentlemen of the Chapel leaving during Elizabethʼs reign, it seems to have been possible at least to retain Catholic sympathies without too much trouble, as long as one was fairly discreet about it, and it seems clear that Tallis was one member of her Chapel who did this.
Certainly his life and music show a very clear involvement with the Catholic party, which, incidentally, was remarkably ubiquitous during Elizabethʼs reign, with very large numbers of the nobility retaining their old religious affiliation. We first see this in Tallisʼs links with the Ropers, a strongly recusant Kentish family—their main family seats were in Kent, at Well Hall (Eltham), Greenwich and Farningham—who were clearly his patrons. Anthony Roper, a grandson of Sir Thomas More (Moreʼs daughter married William Roper, who was also Moreʼs biographer), was a patron of Tallis and a sympathetic Catholic himself. Tallisʼs role as a protector of his pupil William Byrd should not be underplayed, for it was through the good offices of the Roper family that Byrd found somewhere to live in the 1580s, when he acquired from Anthony Roper a house in Harlington near to one of his main patrons, Lord Paget. And, since in about 1591 William Byrdʼs son Christopher married Catherine Moore or More, a great-granddaughter of Sir Thomas More, this Roper link played a major part in Byrdʼs life. It seems to have been through Tallisʼs influence that Byrd gained access to this circle, thus allowing Byrdʼs much more open recusancy to find its expression.
Having explained Tallisʼs recusant links, we have less difficulty in explaining the Latin music he wrote after about 1560, when there could have been little if any opportunity for performance as part of the English liturgy. Wherever else her religious and political sympathies (or duties) lay, Elizabeth clearly enjoyed and supported the music of both Byrd and Tallis: in 1575, Elizabeth issued letters-patent, granting them an exclusive licence to print and publish music. Later that year, they jointly issued what was the first collection of Latin motets ever printed in England, the Cantiones quae ab argumento Sacrae vocantur. Each composer included seventeen motets (no doubt a tribute to this the seventeenth year of Elizabethʼs reign). The motets on this recording all come from this publication.
The Compline Hymn Te lucis ante terminum is one of two settings by Tallis, both identical in structure (verse one, chant; verse two, simple five-part polyphony; verse three, chant): this is the setting based on the Festal Tone, with a middle verse slightly more elaborate in rhythm and texture. O salutaris hostia and Salvator mundi are freely composed, not based on chant or other material. Both are five-part antiphons, the former for Corpus Christi, the latter for The Exaltation of the Cross, and both illustrate Tallisʼs ability to create sustained works from quite short, simple melodic units. The Miserere is a rather different matter, for its ethereal effect disguises a complex and skilful piece of counterpoint. The seven voices contain two canons, one fairly obvious between the top two voices, but another, much more elaborate, within four of the remaining voices: the discantus part is sung by the contratenor in double augmentation (notes four times as long), by the first bassus part in triple augmentation (notes eight times as long) and by the second bassus in augmentation (notes twice as long).
The Lamentations and Spem in alium also fall neatly into the recusant fold: the Lamentations contain the coded references familiar also from motets by Byrd and directed towards recusant circles, while Spem in alium probably has a clear Catholic purpose together with a coded reference of a different kind.
Tallisʼs forty-part work is of course most unusual; there is nothing else in the English repertoire quite like it. The most likely explanation put forward for its composition is that it was written in 1571 in honour of the Duke of Norfolk (a staunch Catholic, executed in 1572), the number of parts presumably referring to years in the wilderness after which the true faith would be restored. The allusion to the restoration of the true (Catholic) faith is found even more clearly in the final admonition of the Lamentations—'Jerusalem, convertere ad Dominum Deum tuum' ('Jerusalem, turn again unto the Lord thy God')—it being a simple and well-understood conceit to compare England with Jerusalem. However, Tallis is not above introducing another conceit of his own, when he chooses a text containing a phrase such as 'Respice humilitatem meam' ('Look upon my humility') and then sets it for forty parts. Tallis lays out his forty parts as eight five-part choirs; the main features which may be audible are an initial passage of 44 beats during which choirs one to four enter, followed by 24 beats covering the introduction of choirs five to eight, an antiphonal passage of 10 beats, and then 10 beats for the full choirs (24 + 10 + 10 = 44 balancing the initial 44); this is followed by 22 beats for 'ne irasceris', during which the stereophonic effect moves back, with choirs eight to five. So at the beginning we have 44, 44, 22; moving to the very end, the extremely dramatic semibreveʼs rest in the middle of the final passage of text (respice humilitatem) provides a final passage of 1 + 32 = 33. There is always much more to a piece of Tudor polyphony than meets the ear!
The preservation of the four-part Mass is another illustration of the influence of highly-placed recusants, in this case the man who commissioned the manuscript which is the only source for this work, Dr Roger Gifford. He was a student and later Fellow at Christ Church and Merton College, Oxford and later (in the 1580s) Physician-in-ordinary to Elizabeth, and it seems likely that in the 1570s he caused to be copied a collection of the music that he had heard in Oxford chapels during the reign of Mary Tudor.
The Mass is composed for four voices, in a rather restrained manner, and it has suffered neglect in comparison with Byrdʼs four-part mass; but the piece is beautifully crafted, contains great variety and deserves to be morewidely known. It is broadly homophonic in style, though we hear the technique Tallis has already perfected in his English-texted music, where one voice leads and the others follow, thus giving a contrapuntal effect very economically, at other times breaking into brief and simple contrapuntal entries. This one-voice-leading technique is also found, of course, at Ierusalem convertere in the Lamentations, where its coming-of-age is established by the powerful effect it conveys through the use of repeated notes.
An understanding of the likely circumstances of performance helps us to address the question of scoring these works. The Mass, being a liturgical work, would normally have been performed in its normal liturgical choral context, transposed up a minor third. However, when Dr Gifford got out his books in the 1570s and 1580s, who knows what forces might have performed it? This question arises more sharply with the Lamentations, which are actually two independent works, but which were associated with each other from an early date (though, strangely, they are often copied in the wrong order in manuscripts, with 'Incipit lamentatio' after 'De lamentatione'). They were only ever conceived for recusant performance, and show the typical terraced-scoring (each part a little higher/lower than the next) seen also in Byrd, for example, which is in contrast to the scoring for specific voices seen in Tallisʼs and Sheppardʼs liturgical music.
Tallis composed during the reigns of four English monarchs—Henry VIII (1509-47), Edward VI (1547-53), Mary Tudor (1553-8) and Elizabeth I (1558-1603)—all of whom had significantly different tastes, both religious and musical. Tallis seems to have responded to each of them with equal ease, with music ranging from the florid early antiphons to simple, syllabic settings: from four to forty voices, this collection of works reveals the remarkably broad compass of Tallisʼs Latin settings.
Roger Bray & Philip Cave © 2000