[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
GG and the harpsichord
Thanks to Dominic Lesnar, here are the liner notes where GG talked about
harpsichords. I'm forwarding it here to the list with Dominic's permission:
-----
You mentioned the liner notes to the boxed set of the Partitas which was
released in 1963. These were reprinted in GlennGould Magazine in the fall
of '98 (Volume 4 / Number 2) as a (spontaneous, off-the-cuff) conversation
between David Johnson and GG.
This is what he had to say about the harpsichord (all italicized comments
are mine [Dominic's]):
----------
DJ: What are your feelings about playing Bach on a piano instead of a
harpsichord?
GG: Let me broach just one aspect of that enormous subject now, an aspect
that fits in with what we've just been considering (they had been
discussing GG's seemingly arbitrary approach to taking repeats, how some
live performers change tempi during repeats, and how GG tries to pull out
inner voices during repeats). There are certain works which you can play
marvelously at one tempo on the harpsichord and you cannot, by any means,
attempt to approach that kind of tempo on the piano with any real success.
People have the impression that the harpsichord doesn't accommodate itself
to quick tempi as well as does the piano, which is made for blockbusting
technicians who want to dash off Liszt etudes and things. But this isn't
true in contrapuntal music. Because of the immediacy of the harpsichord's
attack, because it doesn't know of the sense of legato that is one of the
aspirations of the piano, you cannot attempt the kind of sustaining effect
of the piano- sustaining not in the sense of the pedal suspension but in
the sense of an ideal arched line. You cannot attempt that kind of line on
the harpsichord. You very often find yourself taking the slow movements
somewhat faster than you would ideally like to, I think, and the fast
movements perhaps somewhat slower. The harpsichord offers, depending on
its registration, a certain number of levels of articulation, independent
of anything your fingers are able or not able to do. Whereas the piano is
less circumscribed in that way; there are more possibilities, more choices
to be made.
You know, the piano is not an instrument for which I have any great
love as such. I have played it all my life and it's the best vehicle I
have to express my ideas because it's extremely convenient. It's the only
instrument on which you can fairly suggest an orchestral effect, except,
of course, the organ. I find myself, when confronted with an organ or
harpsichord to practice on, going into wild, Widor-like throes of
improvisation, and this has always been for me a test of how much I'm
enjoying what I'm doing. I very seldom improvise at the piano. I do so on
the organ and harpsichord because of the attraction of being able to
confront an instrument that is registered and still impose an
interpretive will over it. The piano allows so many things to happen that
it presents you with an embarrassment of riches.
The piano is an enormously good instrument for Bach, provided you
approach it in a certain way. I have been greatly tempted to do a Bach
recording on the harpsichord ; it's much easier on the harpsichord- I
think I could do it in one third the time per recording session; I don't
think there would be nearly as many retakes for questions of character
for the reason that the harpsichord does not have the variety of
character in it that the piano does. The piano has so many things that
can go wrong that in order to get those many things to go right, in terms
of inner balances, one must be careful in a subtle way that just isn't
demanded by the harpsichord. This doesn't mean that the piano is a more
faithful instrument to the idea; I am far from happy with the piano as
it's been developed in the twentieth century, and I've done everything
possible in the pianos that I use to castrate them in such a way that
they take on qualities which are almost harpsichordal in tonal
characteristic(remember during Bruno Monsaingeon's Goldberg Variations
outtakes, GG is heard to mention to the piano technician to "think
harpsichord"). The sound I prefer is a very thin one relative to what the
preference is today, relative to the piano as a
Chopin-through-Rachmaninoff vehicle. I have fixed the action in some of
the instruments I play on- and the piano I use for all recordings [CD
318] is now so fixed- so that it is a shallower and more responsive
action that the standard.
(They continue discussing Bach's use of "blocks" as dynamics)
DJ: By "blocks" do you mean a block of forte followed by a block of piano?
GG: Yes, quite. But within that, the major and subsidiary matter can be
varied on the piano as it cannot be on a registered instrument. So the
piano can analyze, but the analysis has to be very acute: you have got to
know what every voice is doing. You can get away with murder on a
harpsichord or an organ, because you can proceed for pages on end just
living on that marvelous clarity which they give you. The moment you sit
down at the harpsichord it says, "I will present your digital qualities as
clearly as any instrument can do for you." Well, the piano does not do
this. It says, "I will make them as muddy as possible unless you present
me with an analytical account so clear that I know exactly what you want."
And everything on the piano, therefore, becomes a very special, inquiring
proposition. I don't mean to be deprecating about the harpsichord, but
everything does not on the harpsichord. It does not impose quite the same
analytical question-and-answer session that you must hold with yourself.
-----
Several comments (me again, not Dominic):
Overall I think GG's comments here make some sense. He makes a decent case
for the need to enliven the textures when playing Bach on the
piano. However, I think he is coming to the harpsichord with some
misconceptions, perhaps inherited from the general thinking about
harpsichord (and the types of instruments available) in the 1940's and 1950's.
First off, GG is a bit off base with his point about not having extremes of
tempo available. True, on a bad harpsichord the tone tends to decay
quickly and this therefore gives some limit to how slowly one can
play. But on a good harpsichord that has some "bloom" to its tone, and/or
with a player who knows how to sustain a line (as opposed to merely
sustaining the sound), slow tempos can be quite effective. I'm thinking
for example of Anneke Uittenbosch's recording of Sweelinck's "Pavana
Lachrimae" (a setting of Dowland's song, "Flow, my tears"), on Globe
5030. She stretches it out to an absolutely incredible 9'49" where a more
normal tempo would bring in the piece in about five minutes, and it works.
And on the fast side, the only real upper limit to speed on either the
harpsichord or the piano is the action's ability to repeat notes. That can
vary as much from harpsichord to harpsichord as it can from piano to
piano. A harpsichord can be played just as fast as a piano, or perhaps
faster since the action is generally lighter and the keys have less
distance to move in their dips.
The factor that most limits high speed on any instrument is: how much
detail does the performer want to put into the individual notes of a
line? If there are sixteen consecutive fast notes and the performer
chooses to give them all equal emphasis, they can be played as fast as
humanly possible, as fast as the fingers are able to move. If instead the
performer wants to give more shape to the line, more levels of emphasis
with some notes being more accented than others, or perhaps with subtle
inequalities of length, the tempo of course has to be slower. This is
particularly important on the harpsichord and organ since there is less
scope for dynamic variation (note, I didn't say there is NO scope for it,
just less). On the clavichord and piano, individual notes can stick out as
louder than the others...indeed they do stick out with any "imperfection"
of finger technique or mishandled hand/arm weight. This can be either a
virtue or a liability. All keyboard players can choose fingerings that
help bring out the musical accentuation of the line. Pianists are
generally trained to make things as even as possible (or to be able to,
even if choosing not to in some musical situations), with all the fingers
delivering notes that sound the same. This was *not* the practice before
Bach's time; quite the opposite. There were "good" and "bad" fingers in
the hand, just as there are "good" and "bad" notes in a melodic line, and
one chooses a fingering that gets the desired sound, using the natural
tendencies of the playing motions to get musical results. Unrelenting
evenness is not such a high goal.
So, if a performer is coming into Bach from 19th century piano technique,
s/he will tend to smooth out the lines and play fast movements rapidly, not
making much articulative distinction among the notes of a phrase other than
perhaps giving some dynamic shape to them. But a performer coming into
Bach from a 16th/17th century keyboard technique will get quite the
opposite result.
Another part of this issue comes from the aesthetic goals of speaking vs
singing. Generalizing here: the 19th century approach is to make the lines
"sing" in long unbroken Wagnerian bursts of tension, many seconds in
duration, and all the notes are subjugated fairly equally under this
"arching line" as GG puts it, the direction of the phrase. By contrast,
Baroque music comes from a rhetorical tradition that is much closer to
speech than to that 19th-century type of singing. Notes are like syllabes,
whether or not there is a sung text. A spoken line naturally has both a
rise and fall to it *and* a huge variety of vowels and consonants and
subordinate accentuations from syllable to syllable. Sure, the line goes
somewhere in the big sense, but the scenery along the way is also
significant. Listen to Baroque opera in Italian, German, or French: the
sung lines declaim the texts dramatically, as heightened speech. The
dynamic and rhythmic nuances (and the basic tempo, and the tempo
fluctuations) come from the way those lines would be spoken. In
instrumental music it's similar: the playing techniques (strings,
keyboards, winds...fingerings, bowings, ornamentation, tongueing, pretty
much everything) described in the historical treatises teach the player to
phrase things as a singer would, with a hierarchy of notes as syllables and
not just a long line pushed through.
An analogy: it's like the difference between driving on an Autobahn vs
driving on a smaller local road. On the Autobahn the goal is to get there
quickly: smooth road, not many exits, focus on destination. On the local
road you also want to get there, sure, but there are more things to look at
along the way, and the road will have more quirks to it thereby limiting
your speed somewhat.
And an example from Bach: look at the bowings for the soloist in Bach's
violin concertos, and in the solo violin and cello works. Or look at the
wind and string articulations in his vocal works. (Use a good clean Urtext
edition, of course, not a 19th-century arrangement.) There is articulative
variety all over the place, from moment to moment, grouping notes in twos
and threes and fours and fives and sevens and whatnot. Sometimes the
groupings are quite unexpected, causing accents across the beats. The
keyboard player can learn a huge amount from studying this music, learning
how Bach habitually thought about the notes, and learning how to invest the
keyboard works with similar variety of expression. Or, the player can
ignore most of that and play long 19th-century lines. It's a choice.
GG was coming to all this with a 19th century view, the default, the way
music was done when he started to form his opinions about it. He's right
that one cannot really attempt a [19th-century type of] line on the
harpsichord. To do so sounds monotonous and rather silly and
unmusical. The notion of terraced dynamics is part of this aesthetic
sensibility: since not much variety is given from note to note, the way to
set apart musical sections is to change registration or dynamic level in
blocks. But there's nothing in a harpsichord's physique that says one
can't do articulative dynamics from note to note, or that a change of
registration is the only way to be expressive! The only thing that limits
dynamic expression is the performer's way of thinking about the music.
GG was of course right that the piano offers phenomenal possibilities of
shading and articulation within a musical texture. And that's what he was
phenomenally good at himself, that simultaneous control of several
things. But I think he failed to realize that the organ and harpsichord
can also do that to a much greater extent than he knew. Sure, it's
possible to bring out separate voices with separate types of articulation,
or to bring out musical motives with articulative emphasis, or to give a
line more legato by holding its notes into the next notes. There are
plenty of things that can go wrong, throwing off the balance in a passage
as GG remarks about the piano. That attention to balance might not be very
important on a bad harpsichord, such as a Wittmayer, but it's quite
important on a good one (especially if real bird quill is used for the
plectra, instead of plastic or leather). Touch is just as important on
harpsichord (or any instrument) as it is on piano: technique is getting a
good sound that is appropriate for the musical expression. Technique is
not some raw ability to deliver fast notes evenly; it's the ability to make
sense of the music.
In the quotations above, GG is right about the modern piano tending to make
textures muddy in ways that the harpsichord and organ do not do. A piano's
tone is not as complex as a harpsichord's or organ's; it's more bland and
generic, and evenly colored from top to bottom of the keyboard
(_relatively_ evenly, compared with a good harpsichord or organ where every
octave on the keyboard has its own character). In short, the piano gives a
more neutral presentation: fundamental tones, not so much in the upper
partials. Therefore within this general muddiness, as GG points out, the
player needs to bring things out more aggressively, and it doesn't happen
as naturally as on the harpsichord or organ.
But that's not to say that the player should be any more lazy on
harpsichord or organ about analyzing the music and playing it clearly! A
boring and generic interpretation is boring, no matter what instrument it's
played on. There exists dull harpsichord playing just as there exists dull
piano playing, or dull guitar playing, or dull flute playing, or dull
singing. One can "get away with murder" on any instrument if the listeners
aren't discerning.
Through all this I think GG simply underestimated the harpsichord's
possibilities. He probably meant well, but he just didn't know the
instrument or its techniques. He recognized some of the basic potential in
that clarity he spoke of in 1963, but he didn't have that in his own
technique (if we can judge by his Handel recording, or those two Bach
excerpts). I suspect that if he had taken the time to work at it,
developing the musical vocabulary to translate his "analytical
question-and-answer" approach into expression on the harpsichord (on a
_good_ harpsichord), the results could have been fantastic. He already had
a great headstart in playing close to the surface of the keys, and in his
finger control. It's a pity he didn't take it further.
Bradley Lehman
Dayton VA
http://www-personal.umich.edu/~bpl