How did you choose the four psalm verses that form Tehillim?
I had a rekindled interest in my own religious background in Judaism starting around 1974.
By 1981, I really had a desire to bring it into my music in some way. I thought that
the most obvious way was to set a text in the original Hebrew, and the most obvious
text to set was the Psalms.
The Psalms are the most musical texts that we have in Judaism. They were written by King David,
who was a great musician of his time. We know they were sung by Levites. I am a Levite — some
of us in Judaism know whether we are a descendent of that Levitical family —
and so I would have been a musician way back when. Therefore, setting the Psalms
was a very natural thing to do, and the added bonus was this:
as opposed to the Torah and the Book of Prophets, which have traditional melodies,
the traditional melodies for singing the Psalms have been lost amongst the Ashkenazim;
that is, the European Jewish traditions. They have been maintained among the Yemenite Jews,
but I’m not a Yemenite and am not that familiar with their tradition of singing. I know of
their tradition of singing and I’ve used some of it in The Cave, but I didn’t grow up with it.
Therefore, it was kind of a green light for me to compose, without having to ignore or
incorporate some kind of pre-existing melody.
That was the basic idea for setting the Psalms. Now, which ones to set? There are 150
of them; very good question. I took the book of Psalms in Hebrew and English, put it on
the piano, and started going through the whole book. My idea was to pick a text that I
could say to anyone, Jew or non-Jew. In other words, it had to be a very universal text,
and the ones I came up with were ones I felt comfortable saying to anyone.
When going through the book of Psalms, were you more drawn to particular textual features
such as repetition of words like “yóm-le-yóm” or striking imagery?
No, the only requirement I had was that I could look someone in the eye, whoever they would be,
and say this. That was my only criterion. Of course the text had to be something that got to
me as well. I knew two of the selections from the Sabbath prayers. The first, The heavens
declare the glory of God, is part of what you say every Sabbath; and Who is the man who desires
life is another part of that same Sabbath prayer. I was familiar with those, so that made them
Is there an underlying structure to the whole work?
What underlies the whole piece and what really makes it work is the use of groups of
twos and threes in a totally free arrangement. This was done entirely by ear, depending
on how I heard the syllables of the Hebrew. Ha-sha-my-im meh-sa-peh-rím ka-vóhd Káil
became 1-2, 1-2-3, 1-2, 1-2-3, 1-2, 1-2-3, 1-2. I had never heard music like that
in my head before.Tehillim was a discovery. I had certainly heard Stravinsky and
particularly the Bulgarian rhythms that Bartók used, which are full of 5/8s and 7/8s,
but my use of changing meters, as a non-stop continuity, began in Tehillim.
Tehillim groups very large measures and some conductors [laughs] are not too happy about that.
There was some discussion, and there is still some thought in my mind, that it might be easier
if there was no bar larger than 7/8. One could re-bar it, there’s no question about that,
but the reason it’s barred the way that it is, is because that’s the way the melodies go.
Groups of twos and threes, changing in all kinds of groupings throughout a piece, became a staple
in many works I’ve done since. Once I did it in Tehillim, it became part of my vocabulary,
and I began to use this in different ways. In Tehillim, everybody is moving more or less
homophonically, except in the canonic sections. It’s used in a totally different context
in the second and fourth movements of The Desert Music, where you have groups in the
brass playing conflicting parts of twos and threes and interlocking together,
a kind of gamelan in changing meters. This creates something in between a repeating
pattern that is always varying and an accompanimental texture. Groups of twos and threes
are also used in Sextet and Triple Quartet. I’m just using it now in Bikini and will
undoubtedly use it in Dolly as well.
Is there a thematic arrangement to Tehillim? You start with Psalm 19, then move to Psalm 34,
back to Psalm 18, and finish with Psalm 150.
The first movement, The heavens declare the glory of God, is basically Abraham looking up
at the sky and intuiting that there’s got to be some intelligence, some consciousness behind
all of this—it all works too perfectly. It’s not just the sun, it’s not just the moon,
it’s the whole universe. There seemed no other way for me to start.
The second movement switches to human character: “Neh-tzór le-shon cháh may-ráh”,
“keep your tongue from evil”, “va-ah-say-tóv”, “and do good”, “ba-káysh sha-lóm”,
“seek peace”, “va-rad-fáy-hu”, “and pursue it”. It’s very, very difficult not to speak
what’s called in Hebrew, le-shon ha-ráh.
the evil tongue. We’re all guilty of it. It’s not a good human character trait and it’s widespread.
Its instances have been detailed very carefully in Judaism because avoiding it
as much as you humanly can is considered a very important moral value.
The second movement is about that and everything associated with it.
What happened with the third movement is [laughs], I was in a situation where the first
two movements were being done by themselves in Germany, at the South German Radio Station
in Stuttgart. Peter Eötvos, the conductor, and I were driving together, and he was basically
asking me, “Are you going to continue in the same tempo?” I knew from the way he put it that
he was saying, “Can’t you give us a break?” [laughs] Respecting Eötvos, I decided to insert
a break and actually stop the music after Parts I and II, and then go on in a slower tempo.
Now, as you may know, I don’t generally write movements. I don’t believe in movements.
I believe that when the music stops, it stops. In any event, I came back and decided
to set the 18th Psalm, with that extremely interesting text [reciting bits of the Hebrew
and English]: Im-chah-síd, tit-chah-sáhd, If righteous G-d treats them righteously,
im-ga-vár ta-mím, ti-ta-máhm, if almost perfect G-d treats them almost perfectly,
im-na-vár, tit-bah-rár, if upright G-d treats them uprightly, va-im-ee-káysh,
but with the perverse, tit-pah-tál, G-d is subtle. Musically, it’s set in parallel
[sings im-chah-síd, tit-chah-sáhd and im-na-vár, tit-bah-rár with parallel melodies].
Actually, the model for that third movement is a soprano and alto duet from the Fourth Cantata of
Johann Sebastian Bach, Christ lag in Todesbanden. There’s a back and forth exchange between
the two voices, as there are in so many of the cantatas, and then a kind of resolution between them.
That was the model for Im-chah-síd, tit-chah-sáhd. The four singers in Tehillim are divided into
pairs and respond to each other. As a matter of fact, the doubling of the four voices with
first the clarinets, then later with oboe and english horn, was also a steal from that same cantata.
Bach, like many great composers before and after him, doubled the voices for support.
It’s an old trick, and it’s one of the best in the book; it makes the singers confident
because they know the pitch is there. If you treat the voices that way, and I do,
and also amplify everything, boosting the oboe and english horn, then you get what
my producer Judy Sherman calls the “voicestrument.” So, when you hear the singers doubled
by clarinets at the end of the first movement of Tehillim, and then that immediately
changes to being doubled by oboe and english horn at the beginning of the second movement,
it’s a completely different vocal timbre.
The last movement, of course, is the 150th Psalm. There’s a coda on Hallelujah, which is the
text that’s been set more than any other in the history of Western music, so that was a challenge.
I really wanted to set it. It’s deliriously overjoyed and it refers to tóf u-ma-chól
drums and winds , which is precisely what I was using in the piece. It was too good to miss.
As a matter of fact, now I’m remembering, I finished that Hallelujah in my studio where
I’m sitting now, here in Vermont. I wrote a lot of the piece here, but most memorably, the ending.
I knew I had it. [laughs]
An important technique you use in Tehillim is canon, an outgrowth of your earlier phasing technique.
Phasing is a word that I coined, but all it really refers to, is a variation of canonic technique.
“Phasing” is simply a canon using a short melodic pattern, as opposed to an extended melody,
where the rhythmic distance between the first voice and the second is flexible and gradually changing.
Piano Phase is a variable canon at the unison. It’s a unison canon where the rhythmic distance between
the first voice and the second voice is flexible. Similarly in Violin Phase and Drumming.
In Tehillim, instead of there being a melodic pattern, there are real full-blown melodies.
Then everyone said, “Oh, those are canons.” But canons are canons. Sometimes the subject
is short and that’s what people hadn’t heard before; that’s why it seemed to be different.
The principle is exactly the same as Sumer is icumen and Row, Row, Row Your Boat,
but instead of having a longer melody you have a short pattern.
There are some long stretches of canon in the first and fourth movements of
Tehillim, and you punctuate the long stretches with harmonic changes.
Would you say you were particularly more sensitive to the vertical aspect in this
work than in previous works?
Well, I had to deal with a piece that was going to be harmonically unified over
a relatively long period of time, so of course I was, but I don’t think I mapped
it out the way I did for Music for 18 from beginning to end like a cycle.
I just worked with the melodic material and then tried to figure out ways to harmonize it.
Ain-óh-mer va-áin de-vah-rím, Without speech and without words, Beh-lí nish-máh ko-láhm,
Nevertheless their voices heard, from the first movement, is a particularly good example of that.
I felt that here I wanted to be able to mirror the text in the music.
What happens is that the whole melody is reduced to just four notes: G, A, D, and E.
Those four notes by themselves are very harmonically ambiguous, and consequently you
find that there are changes of key. So, Without speech and without words/Nevertheless their
voice is heard was something that was open to interpretation, looking at the world around you,
and this is mirrored in this particular section by simply creating a very ambiguous
scale that is capable of harmonic reinterpretation.
Then there is the Hallelujah in D Major. What’s interesting about the ending of Tehillim
is that it ends on the dominant, and it’s a dominant eleventh chord. Like Four Organs,
the tonic is on top and the dominant on bottom, which gives you that sense of the music
still going on even when it ends.
What makes a good ending, in your view? Tehillim sustains such a remarkably tireless energy
that it’s hard to anticipate its end.
Oh, it’s a complete setup. I don’t think there’s a better ending that any one could write for
that piece. A good ending, for me, tries to avoid a V-I cadence.
I was guilty of it at then end of the Four Sections, but alas, I couldn’t think of
anything better to do. Outside of that, there’s no formula for anything.
My early pieces are processes in that when they’re done, they’re done.
But again, those are satisfactory endings because the whole piece is about a process.
Therefore, when the canonic process returns to unison, as in Piano Phase, it’s a perfect
way to end. Tehillim is very traditional Western music key-wise, and in many other respects.
Hallelujah in D major certainly seems like a good way to end.
You said that setting Psalm 150 was a challenge, not because you had to deal with Judaic
tradition but with Western musical tradition. Did you encounter other difficulties while writing
I think I encountered most of the difficulties even before I started composing Tehillim.
When I first put the Psalm texts in front of me, after I figured out which ones I was going to set,
it was as if they were talking to me: “Well, Handel gave me this, Bach gave me that,
Britten gave me this, Stravinsky gave me that - what have you got in mind?”
It was like I was being picked up by the scuff of the neck and asked, “Where’s my melody?”
Once it got going though, it was a complete revelation.
Tehillim is one of those pieces where I did something entirely different in a big way.
When I was first working on it my wife said to me, “You’re actually singing! [laughs]
You’re singing melodies with words!” It was the first time I wrote melodies in that sense.
Tehillim is about melody. It was exciting to do something absolutely different after having
worked for about fifteen years with melodic patterns.
I had been writing melodies in a very short form: - very good melodic module in Piano Phase,
not quite as good in Violin Phase, and very good rhythmic module that lent itself to
very good melodic configurations in Drumming. All those little modules, they have to
be gold or else you’re dead. Melody is always, in some sense, the main element.
There are a lot of melodic patterns in Music for 18 Musicians. They’re short, but they
become extended and work very well melodically. Tehillim is melody in a recognizable way
in Western traditional terms, and that was the break. That really drove the piece and
it ended up being a very inspired piece.
Tehillim took a while to write because it was long, and writing a slow movement was also new for me.
In addition, I hadn’t set a text to music since I was a student, and had never done it successfully
as a student. So, finally I was setting a text that I was excited about, and it was really working.
What the singers were doing in works like Drumming and Music for 18 was doubling an instrument
and using vocalise to become part of the musical ensemble; they weren’t singing words.
Tehillim is the first piece where I said, “Okay, singers are going to sing words and the
instruments are going to accompany them.” Singers love singing the piece and that’s extremely
gratifying. It’s certainly one of the best pieces I’ve ever done.
Why had you avoided using melody?
Basically, I had been interested in getting rid of melody and accompaniment,
the whole homophonic model, and saying, “I am dealing entirely in a contrapuntal situation
of short repeating patterns.” Everything comes out of that web of melodic motives,
not really melodies but melodic patterns. That worked very, very well.
Tehillim was saying, “Let’s see what happens if you just apply that kind of
thing to extended melodies.”
From New York to Vermont: Conversation with Steve Reich
- Rebecca Y. Kim