Tallis was undoubtedly the greatest of the sixteenth century composers; in craftsmanship, versatility and intensity of expression, the sheer uncluttered beauty and drama of his music reach out and speak directly to the listener. It is surprising that hitherto so little of Tallis’s music has been regularly performed and that so much is not satisfactorily published.

This compact disc is the seventh in a series of nine which will cover Tallis's complete surviving output from his five decades of composition and will include the coutrafacta, the secular songs and the instrumental music—much of which is as yet unrecorded. Great attention is being paid to performance detail including pitch, pronunciation and the music's liturgical context. As a result new editions of the music are required for the recordings, many of which will in time be published by the Cantiones Press.

Thomas Tallis Music for Queen Elizabeth
In volume 7 of The Complete Works of Thomas Tallis we meet Thomas Tallis as the composer of Latin-texted motels in the reign of Elizabeth I.

Elizabeth I's reign (1558-1603) was a time of increasing stability in political, artistic and religious life. It was also a time when—despite the fear of Catholic invasion, especially from Spain-England enjoyed growing cultural exchange with continental Europe. In the arts, Elizabeth's reign was a golden age. England's rich, but essentially insular and conservative, pre-Reformation heritage was infused with increasing continental influence, and in both literature and music we see the rise of humanism shaping new ideas and forms. The introduction of printing in England meant that English culture would be more widely disseminated and exported.

Like many of her subjects, Elizabeth I must have had mixed feelings about England's reformed church and the music that accompanied the new liturgy. (Tallis's music for the Anglican liturgy can be heard on volume 6 in this series.) Elizabeth had been brought up as a Catholic by her father, Henry VIII and at times she must surely have hankered after the lost colour and richness of the Sarum Use. Indeed, at the outset of her reign she allowed considerable freedom of practice and belief. Though she was firmly in favour of a vernacular liturgy for the general population, she was happy to license a Latin version of her 1559 Book of Common Prayer, for use in college and university chapels where Latin was understood. In her own chapels she certainly liked a more lavish ceremonial (including choral music) than some of her clergy could stomach.

Elizabeth was the fourth monarch to sit on the throne in Tallis's lifetime. The composer, now in his sixth decade of life, must have regarded the new queen's protestant settlement as another provisional stage in a wearying succession of political and institutional reforms, rather than the start of the status quo we now perceive. Whatever his personal convictions—and the slight evidence we have suggests his sympathies lay with the Catholics—the ever-adaptable Tallis met the challenges of a new liturgy, its new styles and genres, with the imaginative force of a man half his age.

Tallis's pragmatism, though, must have been tempered with sorrow. He had witnessed the wholesale destruction of much of England's church music tradition. Dozens of monastic and collegiate choirs which had cultivated polyphony were now silenced; their choirbooks and partbooks, too, destroyed. He had learned his craft as a church musician in the Sarum liturgy, and served his compositional apprenticeship in genres now defunct—the festal Mass, the votive antiphon. To younger composers such as Byrd, Tallis must have seemed a bridge into an age already receding into folk memory.

Yet if Tallis felt himself the heir to a precious, vanishing tradition, he was not oppressed by the responsibility. In his maturity he brought the virtues of cogency and economy to every genre he cultivated; in his Elizabethan motets he mingled the nostalgic sonorities of the insular, conservative English tradition with the latest imitative techniques from the continent. To the last, Tallis was adaptable and open to the possibilities of new compositional approaches. His restless, questing approach is also revealed in sources which show him revising and reusing his own music.

An important effect of the Reformation was to expose English composers fully to continental innovations. On the continent, under humanist influence and at the hands of Josquin in particular, the motet had developed into a key genre by the end of the fifteenth century, one in which composers could select (or compile) affective or dramatic texts which did not necessarily have a 'proper' liturgical function. Longstanding ways of organising texture—for example, laying out a plainchant in the tenor voice—had been replaced with pervading imitation, with text-based ideas passed around the voices, along with devices such as homophony and contrasts of seating and harmony. Although English composers before the Reformation were clearly aware of continental trends (for example, some of Taverner's antiphons are highly imitative), and indeed were simplifying their styles under both Reformatory pressure and continental influence, they remained loyal to conservative genres and styles well into the 1530s. The years of Reformation, and Elizabeth's protestant settlement, freed the Latin-texted tradition of liturgical propriety, allowing composers to reinvigorate the language and harness it to new, expressive and personal ends.

As a group, Tallis's Elizabethan Latin motets (which number about fifteen) are similar to those of his contemporaries in that they are based on a mixture of liturgical and non-liturgical texts. Absterge Domine and Miserere Domine are both 'devotional' settings (i.e. with non-liturgical prayerful or confessional texts), as is Suscipe quaeso (probably from the Marian period and heard on volume 3). Mihi autem nimis is based on an introit text, whereas all other settings are on texts from the offices. Salvator mundi (two settings), O sacuum couvivium, In manus tuas, In ieiunio et fletu, Derelinquit impius and Spem in alium are all responds. O nata lux de lumine is a hymn text, O salutaris hostia is an antiphon and Laudate Dominum and Domine, quis habitabit are both psalm texts.

The celebrated lamentations (to be heard on volume 8) are similar to many of the works heard on this disc in that they are again motet-style settings of liturgical texts. Their style dates them as Elizabethan and, indeed, there was a brief fashion in England during the later 1560s for setting the Holy Week readings from the Book of Jeremiah.

With the exception of Spem in alium and Miserere nostri, all Tallis's motets are scored for five voices. O salutaris hostia apart, the vocal ranges are more limited than in his pre­Reformation music. The ranges tend to be closer to one octave and a third rather than one octave and a fifth, which raises the possibility that they might have been written for performance at more than one pitch. It is perhaps not too fanciful to suggest that at a lower pitch they could be performed by an ATTBarB combination, perhaps in a private chapel service, whereas if transposed upwards three or four semitones they are performable by an SAATB combination, perhaps domestically where ladies could take the top line.

Some support to this theory is lent by the fact that most of the motets appear in Cantiones Sacrae , the publication of 1575 in which Elizabeth commissioned her two senior Chapel Royal musicians, Tallis and Byrd, to publish 34 motets (17 each) in part book form. Cantiones Sacrae marked Elizabeth's determination that England should be put on the musical map and the intention was almost certainly that sets of part books would be purchased by domestic households as well as finding their way into churches and chapels on the continent. The only motets heard here that do not appear in Cantiones Sacrae are O salutaris hostia, the two psalm motets and (for obvious reasons) Spem in alium. The widening of the 'market' for sacred music to include gifted amateurs may also partly explain the shrinking of the vocal ranges.

While the Latin-texted motets may have been intended for private use, their musical success is demonstrated by the large number of contrafacta that seem to have been made from them. Cathedral musicians in the new Anglican liturgy fitted English words to the motets—sometimes a translation but more often new, unrelated, words. Absterge Domine seems to have been particularly popular since four English versions survive. One of these, Discomfort them O Lord, is heard here whilst most others are to be found on volume 8 in this series. Salvator mundi and O sacrum convivium both survive in two versions each; the other motets with surviving contrafacta are O salutaris hostia, Mihi autem nimis , O sacrum convivium , Salvator mundi II and spem in alium.

Salvator mundi I opens this disc, as indeed it does Tallis and Byrd's printed collection Cantiones Sacrae. The opening section is highly imitative: the five voices enter in sequence from superius to bassus, with arresting rising intervals of fifths and fourths. There is dramatic homophony for the words 'auxiliare nobis' (help us), and a repeated imitative section, 'te deprecantur'. Whilst restricting himself to an economic use of rhythmic and melodic motifs, Tallis nevertheless manages to create a continuous musical unfolding across the entire length of the piece. The text is an antiphon, proper to Matins of the Exaltation of the Cross.

O sacrum convivium appears to have its origins as an instrumental fantasia and then to have been re-written as a vocal piece. An English­texted version, I call and cry, may also date from the mid- 1570s; both texts fit the music equally well, giving rise to doubt about Tallis's original intentions. It is a beautifully crafted motet whose cogency is achieved by means of pervading imitation and climactic sequential repetition. The text is from the Magnificat antiphon of second Vespers at the feast of Corpus Christi.

In nanus tuas is a setting of the respond from Compline. It differs slightly from Tallis's other motet settings in that it has a 22-note compass rather than 19 or 20 and therefore does not lend itself to the possibility of performance at dual pitches. This is not to imply that it was intended for liturgical use; it is still a respond-motet, unperformable in a correct liturgical form.

Tallis's setting of O nata Luxde lumine is similarly 'unliturgical' in structure. It is a setting of the first two verses of the hymn for Lauds on the feast of the Transfiguration. Tallis's liturgical hymn settings (heard on volumes 4 and 5) all begin in triple time and Tallis honours this tradition in this motet setting. Despite its near­continuous homophony the work is a gem: phrase lengths are cleverly varied, modulation is swift and well-planned, and the occasional inner part motion is motivically cogent. Tallis indicates a repeat of the last line of music which is unique in his motet settings, but reminiscent of his English anthems set in ABB form.

Absterge Domine was evidently one of Tallis's most popular and enduring motet settings. Not only does it appear in Cantiones Sacrae but it also survives in no less than four contrafacta—later reinvented versions with differing English texts. The extended 'confessional' text results in one of Tallis's longer settings and, unlike in some of his shorter motets, Tallis mirrors the textual punctuation in the music, resulting in a series of clearly defined musical phrases.

Discomfort them O lord is one of the four surviving contrafacta based on Absterge Domine . The scribe undertaking the adaptation sticks faithfully to Tallis's notes, subdividing and altering durations as necessary to fit the new syllables to the existing notes. As an illustration of our proposed dual pitch theory, Absterge Domine is here performed at low pitch and Discomfort them O Lord at high pitch.

Domine, quis habitabit and Laudate Dominum are Tallis's two surviving settings of Latin psalm texts. They are part of an extensive Elizabethan tradition of psalm settings, to which composers such as Christopher Tye, Robert White and Tallis's younger contemporary William Mundy also contributed. These two substantial motets have imitative sections alternating with homophony. Domine, quis habitabit is a setting of psalm 15, and the rather more successful Laudate Dominum is a lively selling of psalm 117 (only two verses long) and is notable in that it includes the Gloria Patri.

Miserere nostri follows a continental tradition of complex canonic demonstrations of technical skill; it is partnered in the Cantiones Sacrae by Byrd's double canon Miserere mihi Domine. The setting is a canon six in two; six voices are used to create a simultaneous or double canon. The first is a canon at the unison, between the two highest voices. Superius I is the antecedent—the first sounding voice—by one semibreve. The second is a mensuration canon of four voices, au beginning simultaneously. The Discantus part is the antecedent with the Contra tenor in canon in double augmentation, meaning that the notes are four times longer. The two Bassus parts are in canon 'per Arsin et Thesin', that is, they are inverted so that upward intervals in the antecedent are downward in the consequent (the answer). Bassus 2 is augmented—so that the note values are doubled—and Bassus 1 is triple augmented; the note values are eight times longer. The seventh voice is a 'free' tenor part, though only in one place is its presence required to complete the harmonies.

Like O sacrum convivum, Tallis' second setting of Salvator mundi may also have begun life as an instrumental fantasia. It may have been conceived as a two-part canon (at the octave) with a bass part and may have had its first five­voice incarnation as the English-texted version When Jesus went into Simon the Pharisee's house. If this is the case the Latin-texted version represents a further and final revision, in which the non-canonic voices were substantially recomposed.

Mihi autem nimis sets the opening text of the introit for Mass on the feasts of the Apostles. Tallis approaches it in devotional mood and creates an intimate and finely wrought setting which, like Salvator mundi I, forms a continuous phrase of music from start to finish with little audible punctuation.

O salutaris hostia stands apart from Tallis's other motets in two ways. Firstly it appears to have existed in several revisions during the course of its lifetime and (especially since it does not appear in Cantiones Sacrae ) the modern editor has a number of decisions to make in producing a single 'definitive' version. Secondly, the five voices are widely and evenly spaced, more akin to the SATBarB arrangement of pre-Reformation music than Tallis's usual pattern of employing similar voices in the second and third parts from the top. The text is the fifth verse of the hymn at Lauds at Corpus Christi, but is probably more familiar as the opening of the hymn at Benediction.

In ieiunio et fletu and Derelinquit impius are almost certainly among the last works that Tallis composed. They are markedly experimental settings of Lenten, penitential texts which can be read as especially apposite to the plight of the recusant Catholic community. In ieiunio et fletu tells of weeping priests who beg to save their heritage from destruction whilst Derelinquit impius is a plea for the sinful to return to the Lord. In the second work Tallis was clearly preoccupied with the expressive possibilities of modulation and of denying the gravitational pull of a 'a concern signalled at the outset by an imitative exposition in which voices enter on unexpected degrees of the scale. The work is harmonically conceived, with much of its interest achieved by chromatic means. In ieiunio et fletu takes this tendency further, dispensing with standard imitative techniques altogether and replacing them with canon and repeating blocks of texture; in addition the nominal 'tonal centre', G, is not established until the closing bars, thereby creating a disorientating aural effect. Both motets take their texts from the Tridentine Matins on the first Sunday of lent. They are the third and fifth responds respectively, and it is surely no coincidence that in Cantiones Sacrae they appear with William Byrd's setting of the fourth respond, Emendemus in melius. Perhaps the two composers intended them for use by the recusant Catholic community.

A further characteristic of In ieiunio et fletu is that Tallis scored it very low. In order to perform the motet with the usual combination of voices, upwards transposition of more than half an octave is required. Performance at the scored pitch makes for very sonorous and rich textures but requires a bass who can sing a low D! For comparative purposes we have chosen to perform In ieiunio et fletu at both low and high pitches on adjacent tracks.

Spem in alium is surely not just the greatest of all Thomas Tallis's musical achievements, but one of the great musical compositions of all time. Written for 40 independent voices, this is a noble and monumental edifice which in the course of its 69 longs makes creative and imaginative use of the extensive musical palette.

Tallis groups his singers into eight choirs of five voices (soprano, alto, tenor, baritone and bass) and it seems most likely that he intended them to stand in a horseshoe shape. The piece begins with a single voice from the first choir; gradually the voices enter in imitation and, as the earlier voices fall silent, the sound moves around the line from choir one to choir eight. During the fortieth breve, all forty voices enter simultaneously for a few bars, and then the process happens in reverse with the sound moving back from choir eight to choir one. After another brief full section the choirs sing in pairs alternately throwing the sound across the space between them until finally all voices join for a full culmination to the work.

Clearly Spem in alium is an occasional piece despite being based on a liturgical text; 'Spem in alium' is a respond from Sunday Matins during the reading of the history of Judith. Various theories have been put forward concerning the purpose for which Spem in alium was written and the significance of the number of voices. Of these, Paul Doe's suggestion that the first performance took place in 1573, the fortieth year of Elizabeth l's reign, was originally the most plausible explanation.

However, as Denis Stevens Iater pointed out, a near contemporary account from 1611 describes how Tallis was commissioned to compose the work—probably by Thomas Howard, fourth duke of Norfolk—as an answer to Striggio's 40- part Ecce beatam lucem . This may place the first performance in the long gallery at Arundel House on the Strand, perhaps in 1570, after Norfolk was released from prison (he was executed in 1572). It is intriguing to note, too, that the banqueting hall of Nonsuch Palace—Norfolk's country home—was octagonal and possessed first-floor balconies.

The earliest surviving manuscript of this great work, the Egerton manuscript, is laid out with an English contrafactum, Sing and glorify heavens high majesty. This version was evidently produced for the coronation (as Prince of Wales) of Prince Harry in 1610 and (after his untimely death) repeated in 1612 at Charles' coronation. In the manuscript Harry's name is clearly written in each part—then crossed out and Charles' name substituted. The English words are not a translation of the Latin, but a new poem written as a syllable-for-syllable replacement. Evidently the authorities decided that musically Spem in alium was fitting for such an impressive occasion as a coronation, but that the Latin words were too somber.

An interesting feature of Spem in alium is that its total length is 69 longs (a long being two breves). This is a cryptogram; the same number is arrived at by taking Tallis' name, ascribing each of the letters of the Latin alphabet a number (A = 1, B = 2 etc.) and summing the values. Perhaps it is not too fanciful to imagine that Tallis 'signed' the work in a way that ensures he is fully bound up with his summa for perpetuity?

Alistair Dixon © 2000