Brett Mitchell on Gershwin’s Piano Concerto in F

Skype Interview 15th October 2018


Hailed for presenting engaging, in-depth explorations of thoughtfully curated programs, Brett Mitchell began his tenure as Music Director of the Colorado Symphony in July 2017. Prior to this appointment, he served as the orchestra’s Music Director Designate during the 2016-17 season. He leads the orchestra in ten classical subscription weeks per season as well as a wide variety of special programs featuring such guest artists as Renée Fleming, Yo-Yo Ma, and Itzhak Perlman.


Mr. Mitchell is also in consistent demand as a guest conductor. Highlights of his 2018-19 season include subscription debuts with the Minnesota Orchestra and Vancouver Symphony Orchestra, and return appearances with the orchestras of Cleveland, Dallas, and Indianapolis. Other upcoming and recent guest engagements include the Detroit, Houston, Milwaukee, National, Oregon, and San Antonio symphonies, the Grant Park Festival Orchestra, the Rochester Philharmonic, the Saint Paul Chamber Orchestra, and the New Zealand Symphony Orchestra. Mr. Mitchell also regularly collaborates with the world’s leading soloists, including Yo-Yo Ma, Renée Fleming, Rudolf Buchbinder, Kirill Gerstein, James Ehnes, Augustin Hadelich, Leila Josefowicz, and Alisa Weilerstein.


From 2013 to 2017, Mr. Mitchell served on the conducting staff of The Cleveland Orchestra. He joined the orchestra as Assistant Conductor in 2013, and was promoted to Associate Conductor in 2015, becoming the first person to hold that title in over three decades and only the fifth in the orchestra’s hundred-year history. In these roles, he led the orchestra in several dozen concerts each season at Severance Hall, Blossom Music Center, and on tour.


From 2007 to 2011, Mr. Mitchell led over one hundred performances as Assistant Conductor of the Houston Symphony. He also held Assistant Conductor posts with the Orchestre National de France, where he worked under Kurt Masur from 2006 to 2009, and the Castleton Festival, where he worked under Lorin Maazel in 2009 and 2010. In 2015, Mr. Mitchell completed a highly successful five-year appointment as Music Director of the Saginaw Bay Symphony Orchestra, where an increased focus on locally relevant programming and community collaborations resulted in record attendance throughout his tenure.


Born in Seattle in 1979, Mr. Mitchell holds degrees in conducting from the University of Texas at Austin and composition from Western Washington University, which selected him in as its Young Alumnus of the Year in 2014. He also studied at the National Conducting Institute, and was selected by Kurt Masur as a recipient of the inaugural American Friends of the Mendelssohn Foundation Scholarship. Mr. Mitchell was also one of five recipients of the League of American Orchestras’ American Conducting Fellowship from 2007 to 2010.


Well, there’s a really interesting mix of things here to discuss! It was very recently you conducted this, wasn’t it?


Yeah, it was on September 28th, 29th and 30th – two weeks ago. So it’s really fresh for me.


And who was the pianist?


It was Joyce Yang.


So when you programme a concerto that’s less commonly performed (like this one), how does it come about? Does the pianist or the conductor/orchestra suggest it?


Well, it really depends on the project, and it depends on the artist. Look, I love a great Mozart overture, Beethoven concerto or Schubert symphony. But if they don’t have anything to do with each other, sometimes it feels like you’ve just slapped three pieces onto a programme. I was given a really great piece of advice very early on in my career. I was talking to one of my mentors about programming, and he said ‘try thinking about it like this: if the three pieces on this programme all sidled up next to each other at a bar, what would they have to talk about?’


I love that.


I really like that, because if you have three pieces that have nothing at all to do with each other, then it’s very difficult to have a conversation (if you take my meaning). If you have three pieces that are too closely related that can become difficult too. So what I was trying to do with the Gershwin concerto was for it to be on a ‘jazz in the concert hall’ type programme. We opened up with a really fantastic piece by Duke Ellington called Three Black Kings (the last piece he wrote before he died), then we went into the Gershwin Concerto in F. In the second half of the programme we did a piece by John Adams (who’s not normally associated with jazz) called City Noir (2009). It’s kind of a jazz symphony in all but name. In this particular case, we really knew that we wanted to have the Gershwin Concerto in F on this programme, so it was a question of who were the possible candidates for it. Our artistic team at the Colorado Symphony did a bit of digging around and ringing up managers, and we actually settled quite quickly on Joyce. She was just amazing – could not have been a better fit for the programme.


Brilliant. So how did you prepare to work with her? Were you in communication about the details of the music in advance or was it more just during rehearsal?


It was, but it really depends. The Gershwin Concerto in F is not one of the top thirty standard repertoire concerti that get played, but it certainly gets played a good amount over here in the States. The orchestra knows it. Joyce (obviously) knew it. Certainly, in my wheelhouse, I was in love with jazz long before I was ever in love with classical music. Gershwin, Bernstein and people like that were a great bridge for me. So unless its something that’s really ‘out there’ and it’s a composer that I (or the orchestra) haven’t done a lot, we don’t sort out that much ahead of time. Joyce and I just met like I would any other soloist beforehand; we didn’t really have any communication before she got to town.


Our normal rehearsal schedules are that we do a single on Wednesday, a double on Thursday, and then the dress rehearsal on Friday before we do the shows on Friday, Saturday and Sunday. On Thursday mornings I work with just the orchestra on the orchestral portions of the programme. Then over our lunch break (between the two rehearsals) I sit down and meet with the soloist. That’s exactly what Joyce and I did. The Thursday afternoon rehearsal (from 1pm to 3pm) was our main rehearsal with Joyce before we did the dress rehearsal the next morning.


I see. So it’s always a very small time frame.


Yeah. I think we get so accustomed to the schedule, but it’s funny if you really think about it. I meet soloists for the first time at noon on Thursday and then thirty-one and a half hours later, at 7.30pm on Friday evening, we go on stage together for the first performance. It’s like speed dating; but with two of you, plus the hundred people in the orchestra and the two thousand people in the audience. It’s a very public and quick marriage, as it were.


Haha! Arranged.




Obviously it’s a very big showpiece. How large is the string section etc.?


A lot of what that gets determined by, of course, is the size of the wind and brass sections. So in this particular piece it’s all winds in threes (except for the bassoons), four horns, three trumpets, three trombones and a tuba, along with the percussion. We’re talking a full symphony orchestra here, so you really do need your full strength of string section. It’s not like doing a Mozart or Beethoven piano concerto, or Vivaldi’s Four Seasons where you really cut down on the string numbers (in Vivaldi’s Four Seasons there are a couple of winds that play). With a big full section like this, we had our full string compliment out there. So I think that was something like 14, 12, 10, 9, 8.


Right. And three percussion?


It is timpani plus three additional percussion. So it really is a full orchestra.


Yes it’s big.[1]


There were quite a few cuts made to this piece when it was premiered, weren’t there?[2] When I spoke to Leonard Slatkin about Copland’s Third Symphony, Slatkin said he was the driving force behind the music that was cut by Bernstein being reinstated.




– so now it’s quite well known what had been missing. Is that the case with Gershwin’s piece?


I don’t really think so. It’s funny you should mention that, because two weeks before we did the Gershwin, we did the Rachmaninov Symphony No. 2 (on our opening weekend of our classical season here in Denver). Now that piece, of course, is notorious for having been cut and cut. I think an uncut performance lasts around an hour but at certain times it was cut down to something like thirty-five minutes. You’re talking about eliminating almost half the piece! It hasn’t been recorded in a cut version for almost fifty years, I think. Almost every performance that you hear of the Rachmaninov now is performed complete (with the exception that most people usually omit the repeat in the first movement).


As you can imagine, I do Copland’s Third Symphony quite a lot as an American conductor. I did three performances of it with the Cleveland Orchestra about a year and a half ago. That was the first time I had ever done the uncut version. Then I did it again later that same year with the National Repertory Orchestra in Breckenridge, Colorado, and twice that November with the San Antonio Symphony down in Texas. My undergraduate degree was in composition – I take what composers say very literally. I always feel that if a composer wrote it then we should play it. But I have to say (and I’m going to have to disagree with my compatriot Leonard Slatkin here, who was a mentor of mine) I feel that maybe Bernstein was right, because the end of that Copland –


– I know.


It’s like – how much triumph do you want?! We get it, you’re happy! Yes, the message has gotten through!! So, when I did it again in Breckenridge over the summer and San Antonio in the fall, I actually restored those Bernstein cuts. I think the piece is stronger for it, and that’s coming as a recovering composer myself so I don’t say things like that lightly.


It’s an interesting idea about cutting the Gershwin. I’m sure that was done at some points. But I’ve never thought of cutting it. Although it’s unusual in some ways, structurally, if you take out certain sections (even if they do feel a little redundant) to me the proportions get off a little bit. So I always do it uncut.


Ah I see. I thought I would check, and actually in Copland’s case, he came round to Bernstein’s changes anyway.


Yes, and the only way it was ever published was with those cuts, so you couldn’t even know what the cut music was. When Marin Alsop, (who was also one of my predecessors here at the Colorado Symphony) would restore the music that had been cut, my understanding is that it was all handwritten because it had never been published. In fact, it wasn’t until Boosey and Hawkes published it maybe only two years ago. You can’t even purchase it. So it’s a very interesting question, ‘do you cut or don’t you cut?’


Gershwin doesn’t put in that many metronome markings, does he? You have certain ones at pivotal points but a lot of it is left to you. How much of that is set/suggested by you and how much by the pianist?


You know, that’s a great question. There was a television show on in the Eighties and Nineties called ‘Who’s the Boss?’ – and that’s how it always feels doing a concerto. And my approach is (and this is my overall approach to conducting): I’m there as a facilitator. I didn’t create any of these notes – that was the composer. I don’t execute any of these notes – that’s the musicians. I want the soloist to sound as great and feel as comfortable as they possibly can. I am there to serve the composer, the soloist, the orchestra and thereby the audience. What that means is that sometimes, when we go into that Thursday afternoon meeting, I have to have a radical reimagining on the spot.


Maybe the soloist wants to do half note=88 instead of half note=69. That’s a very different kind of feel. I try to get inside the mind of the soloist, to figure out what it is that they’re trying to express through their performance. As long as we can get on the same page, I’m very happy to go with the soloist. Maybe that’s just being a child of the Eighties, I just take a collaborative approach to what we do. Technically I am ‘the boss’, but at the end of the day I literally cannot do what I do without the orchestra, the composer and the soloist.


A short anecdote. One of my mentors (who shall remain nameless) was doing a performance when he was a student at Tanglewood. He was conducting an act from La Boheme. He told me a story of what the soprano said to him in the middle of a rehearsal: ‘actually Maestro I take a little bit of time here’. He said ‘oh that’s nice. I don’t’. While that’s an amusing story, there may well have been a very good reason why the soprano was taking time there. Maybe she needs a breath, maybe she is trying to express something differently in the phrase, so the idea of ‘my way or the highway’ just doesn’t accomplish anything. You’re now disadvantaging your soloist and putting them in an awkward place. I guarantee you, no matter what the concerto is, what the soloist has to do is way more difficult than what I have to do. Period. So then you can get to a place where you can realise that your job is to empower this person in their interpretation as much as you possibly can.


Are there any specific moments when you’re more likely to deviate from the tempi in the score, or alternatively more likely to stick to them?


There’s not anything specific that comes to mind but I will have conversations with soloists where I have to say (and this tends to happen with younger soloists who are a little bit less experienced, through no fault of their own): ‘I know you can play it that fast but the trombones can’t, it’s just a physical impossibility. Could we maybe step off the gas just a little bit here so that everybody has the opportunity to articulate the phrase in a way that’s comprehensible to the audience?’ But you can hear the language I use. I try to make it as collaborative as possible, rather than ‘why is this so fast? This cannot be this fast. The trombones can’t do that. This is ridiculous!’ That’s not the way to get anything done. I have that conversation a handful of times every season. Ordinarily every soloist says ‘yes of course, my mistake. I get a little excited when that part rolls around’. I say, ‘get as excited as you like just reign in on the tempo’!


Ah OK! What about rubato then? In this piece there’s so much of that. How does that work with the pianist?


Well, I’ll give you an example: the second theme in the first movement at R4. It’s the piano’s first entrance by itself. That theme gets restated a whole bunch of times throughout the course of the movement. What I’m really trying to do in a rehearsal with a soloist and an orchestra is what I would call ‘directed listening’. When the orchestra rests from R4-5, they should not just be sitting there. They should be listening to the phrasing and the rubato just like I am. It makes zero sense if the pianist plays this second theme and speeds up through the first two bars (with the crescendo) and then pulls back the tempo through the second two bars (with the diminuendo), and the orchestra and I don’t listen to that. It’s really about directing the orchestra’s listening. Sure, I can (and do) show it, but I would much rather the orchestra take ownership over the music making. I’m happy to help steer them in the right direction, and to make sure we have uniformity of purpose. In my experience, that’s how you get the rubato – by making sure that the players always know that they have to be actively listening and not just playing along.


You said you did three concerts with Joyce – does her use of rubato vary from performance to performance?


I think it does. Whether you intend for it to or not, we’re all humans, and there’s eighty of us up there on stage. Maybe Joyce had an extra scoop of oatmeal that morning, so the tempi are a little bit slower. Or she had an extra shot of expresso in her Americano, so her tempos are a little faster! Again, that’s why I so heavily encourage that directed listening, because I don’t ever want any of us to go on stage feeling we have to do it tonight the way we did it last night. First of all, it’s never going to happen, and that’s good – they don’t want to play it the same way three times.


I try to be very flexible, within reason of course. You’re not talking about if yesterday you played it quarter =104, and then at quarter = 80 today. That’s silly, now you’re changing all sorts of things. But if the tempo’s a touch slower or faster today, maybe instead of pushing forward through the crescendo we actually stretch the crescendo, so the arrival of it really feels like an arrival. If you just say ‘Joyce is going to take some time in the third bar of R7’, you’ve set it up so that Joyce has no flexibility and it also in some ways allows the orchestra to check out. What if Joyce doesn’t feel like it today? And what if, with an orchestral piece, I don’t feel like that today? What if I feel like maybe we overdid it or underdid it last night? Again, it’s the flexibility of approach that allows for the flexibility of tempo.


Ah that’s really interesting. Not long ago I spoke to three conductors about dance, and actually it’s not quite the same dynamic. If a ballet dancer is taking their time over a certain step, obviously the orchestra can’t see that, so it will be the conductor they rely on.


I’ll piggy back on that. About two years ago I did seven or eight performances of The Nutcracker with the Cleveland Orchestra – we brought over the Pennsylvania Ballet from Philadelphia. Doing seven performances, you have a different cast, and the Sugar Plum Fairies ranged in height from 5’3” to my height, 6’1”. Obviously for the 6’1” Sugar Plum we had to slow it down. Those limbs are longer – it takes longer to take each of those steps! So there is that flexibility as well, but in a different way. You’re exactly right.


Yes, conductor Gavin Sutherland said that the ‘White Swan’ duet in Swan Lake can range from seven to twelve minutes, something ridiculous like that![3]


Now to the synchronisation of the actual music. There are lots of places (like p.97 or p.103) where the music in the orchestra is fairly straightforward but the piano part is relentless semiquavers. Is that tricky?


You know, sometimes it has to be rehearsed, but generally speaking the pianist (or any instrumental soloist) tends to understand when their role is accompanimental. If you look at p.97 at R13, the tune is clearly in the strings – it’s from the second movement R10 and comes back here at R13 in the third movement. So it really is incumbent upon the pianist to make sure that he or she is keeping it steady and at the same time keeping that tune in their ears. We’re not going to chase around the pianist playing the semiquavers. I mean, you can see it on the page: nothing about that screams out ‘rubato flessibile’. It’s the same at R19 on p.102-103 (that comes from part I in the second movement). These are all recalls from previous movements, the pianist really has to go with the orchestra there.


In terms of synchronisation, passages like that are actually not as difficult as you might think. It’s the corners, the angles, and the transitions that are actually difficult. That’s the only time that the question ‘who’s the boss?’ really comes into play. For example, when we have to meet up at a certain downbeat, we have to know if the orchestra really needs a nice firm, crisp, clean, clear downbeat there. If they do, what that means is that I probably need to dictate that downbeat to the orchestra and the pianist as well. There are other times where the pianist can dictate exactly where that downbeat falls.


It’s entirely dependent on the character of the music, on what the orchestra visually needs from me, and who’s got the tune. If the strings have the tune and the pianist is accompanying, the pianists in particular tend to be pretty good at this. I used to be a pianist, so I know. We tend to be quite good at accompanying. So the vast majority of soloists are completely understanding of that. Some soloists that I’ll work with will say ‘you know what? It’s probably just easier if you give that downbeat there and then it will be sure of being together’. It depends on the soloist, the piece, the context, what the orchestra needs, but those transitional moments are the only time when that ‘who’s the boss?’ question actually does come up.


For example, back at the very beginning of the piece if you look at p.10 R7, the pianist finishes out that that second theme at R6 at the top of the page, and then there’s a little tag Molto meno mosso.


That bit is great.


It’s wonderful, but it depends on how much rubato the pianist wants to incorporate. Some players go [sings a steady pace], and that’s quite easy to follow. Some pianists go [sings a slow start, speeding up] and I just think ‘well I’m hoping I’m counting to sixty-four correctly!’ because that’s the number of notes between the molto meno mosso and R7. It’s angles and corners like that get discussed during that meeting that we talked about (just me and the soloist before we get together with the orchestra), so that I know exactly what the soloist’s intention is. But it doesn’t mean we get it together the first time exactly in rehearsal.


Compounding the difficulty of getting into R7 is that C major chord. The right hand part for the piano part is no problem, but the left hand has that C and E that’s a tenth apart, and many pianists cannot reach that tenth. What that means is you end up with that low C sounding before the E. So, are we all coming in on the downbeat with the low C? Or are we coming in when the chord happens? You add that after all of the rubato in the sixteenth notes and the subito molto rit. that happens beforehand – those are the kind of angles that are challenging. If you just flip the page going into R9, that’s no problem at all, because we’re just going into a new tempo. I don’t worry about that because I have four bars to catch exactly what the tempo is before I have to help the orchestra in the fifth bar of R9 (where the first flute, piccolo and second clarinet are doing their little scale up to the downbeat of the bar there).


Two bars before R9 you’ve got the flutes, clarinet, bassoons, all four horns and all of the violins and cellos. That makes a reasonable amount of noise. So when you get to R9, if the pianist wants to start that phrase quietly that’s a little difficult to catch after a loud passage. So it’s good news that you have those four bars. Every little transition is its own puzzle to solve, and it’s a puzzle that two of us (the conductor and soloist) try to figure out how to approach in our meeting. Then we see how effectively it works in rehearsal. With R9, we didn’t have to go over it more than once, but at R7 we had to get into that passage maybe two or three times. Again, even though they’re taking it from my hands I’m still always trying to direct the orchestra’s listening. I’m trying to encourage them to let the pianist show them exactly how they’re trying to get into R7 so that they can take some ownership of it too, rather than just do as I show at the exact moment I show it. There are some places where that’s necessary, but I try to keep it as collaborative as possible. Those transitional corners, they’re the challenge.


Ah right. It’s a very transitional piece, generally, isn’t it?


Oh yes.


I read that Prokofiev wasn’t keen on Gershwin at all,[4] partly because he thought his music ‘not much more than “a succession of thirty-two-bar choruses ineptly bridged together’ (cited Greenberg 1998:140).




Gershwin was very keen on using effects on brass and piano, are there any difficult passages where you have to consider the piano in that way?


Not really to be honest, because he’s so clear. Gershwin is as clear with that kind of thing as composers like Stravinsky or Ravel. There are all these great, magical, deeply interesting colours, but short of just making sure they’re audible to the audience, there’s not really much that I have to do to make that happen. Most of it is just encouraging those interesting colours to come out.


Ah, so Gershwin doesn’t really demand a different mindset from you.


That’s right.


Quite a few people that Gershwin approached for lessons turned him down, didn’t they? Nadia Boulanger, most famously, because she thought her very academic approach would stifle his instincts.[5]


Yes, that’s exactly right.


And there was Prokofiev (who wasn’t a fan) and Ravel, who turned him down and famously said ‘why be a second rate Ravel when you can be a first rate Gershwin?’ I love that quote.


I do too.


You could apply that to anything in life, couldn’t you?


Exactly. I get asked all the time if I listen to recordings when I’m preparing to conduct whatever it is I’m about to conduct next. My answer is that I’m happy to listen to them at the very end of the study process. I went to school, and continue paying pack my student loans, just to make sure that I could read these scores every bit as well as anybody else. So when I’m learning or getting reacquainted with a piece, I do that by myself. I’m not interested in having another conductor’s interpretation influence me. At the very end of the process I’m happy to listen to five or six different recordings, just to see what the trends are. I’ll listen to a recording from perhaps the forties, the sixties, the eighties and then something contemporary.


The reason I bring that up is because my standard answer to that has always been: ‘I would rather be a first rate Brett Mitchell than a second rate Leonard Bernstein’. Leonard Bernstein already did it! He already had the insights to the piece that he had, I will have whatever insights to the piece that I have. It does me no good to try and replicate whatever it is I think Bernstein was trying to bring out of a particular piece. He’s Leonard Bernstein, and I’m not. I’m Brett Mitchell, and I don’t want anybody trying to do as I do either. If you don’t have something to say then stop speaking, it’s pretty straightforward! I’m sure it was disappointing for him, but when somebody like Gershwin gets turned down by Boulanger or Ravel, I think that’s the greatest compliment that you could possibly receive. There’s an alternate version of the Ravel story. Which goes: ‘Mr Ravel I’d like to study with you’. Ravel asked Gershwin how much money he made last year, and Gershwin told him. Ravel answered ‘then perhaps, Mr Gershwin, I should study with you’.


Exactly! Yes I’ve heard that one too.


I think that’s exactly right. If you look at both of the Ravel piano concertos from a few years later, they sound remarkably like Gershwin! It’s not like Gershwin was lacking in craft. And listen Prokofiev, with his thirty-two bar problem, whatever. I would encourage him to go back and listen to Schubert songs, most of which are thirty-two bars. It’s not the form that’s the problem. Perhaps it was the language, perhaps it was the tunes, perhaps it was the success. Who knows? But I’m glad that Gershwin was Gershwin and sounds like Gershwin, I don’t want him to sound like anybody else. I tell young conductors that all the time – when you get on a podium in front of an orchestra, the easiest thing is to put on airs, and to try to present yourself as something other than who you are, because you’re not secure with who you are. But musicians don’t respond to that, they respond to somebody that is genuine. I think the reason that musicians and audiences respond to Gershwin is not because it’s jazzy and accessible, it’s because he’s legitimately being who he is. George Gershwin wasn’t trying to be anybody else. I’m glad he wasn’t and I love him for it.


Exactly. I find it interesting that Nadia Boulanger just knew her very academic approach would stifle Gershwin’s more instinctive way of composing. Do you think that his music demands a different approach from you? Do you find yourself conducting in a more instinctive way, or have a less academic approach when you’re studying the score?


You know, I don’t actually – it’s a great question. It’s funny, but I don’t view them any differently. If you go through my Concerto in F score versus the one I did two weeks before, Mozart’s K503, you’ll find the exact same kind of markings. There’s phrasal analysis, formal analysis, harmonic analysis… These are the building blocks of music whether it’s Mozart in the late Eighteenth Century or Gershwin in the early Twentieth Century. To some extent music is always music, and there’s always going to be phrases, themes and structure to be contended with – that’s my job. Players tend to think a bit more micro, just because of the nature of what they’re doing, and conductors think more macro. If the trumpet player has a Bb the trumpet player is either going to play a Bb or not, and there’s nothing I can do to make that Bb happen. What I’m supposed to be doing is making sure that if that Bb actually speaks then it makes sense in the context of everything else around it. So whether that’s Gershwin or Mozart, it really doesn’t matter to me – the goal is always the same.


I would also say that it’s the exact same question when you’re talking about directed listening and rubato. There’s obviously a good amount of rubato in Mozart as well, it just happens to be done a little bit more subtly. Mozart wasn’t writing ritards or allegro molto and then eight bars later largo and poccissimo meno. But music is always music and the phrases are always phrases – so no, I don’t think it requires a different approach.


I would consider myself a polylinguist, musically speaking. I speak jazz, it was one of my very first great loves, and obviously I now also speak orchestral/classical music or whatever we want to call it (nobody has any idea what we’re supposed to call it anymore, but you know what I’m talking about!). Whether I’m conducting Duke Ellington, George Gershwin, John Adams, Bernstein or Rachmaninov – there’s phrasing in all of that. The grammar and the syntax are the same you just have to learn how to translate it – just like when you’re learning to speak a different language.


Talking of jazz – did you see the performance on YouTube with the Berlin Philharmonic?[6]


No, I didn’t.


It wasn’t what I expected, they performed it with the Marcus Roberts trio and Seiji Ozawa was conducting. Instead of being about a 32-35 minute performance it lasted nearly 45 minutes – with each cadenza the trio were left to improvise. Do you hear it done that way often?


Well, I remember hearing about that a few years ago. I think it’s great. I was in a jazz trio of my own, I was the pianist and two of my best friends were the bassist and the drummer. So I love all of that and I think it’s a wonderful, interesting, innovative approach to doing the piece. I’ve never done it – I would be totally open to doing though! Obviously assuming the Gershwin estate was OK with it, which it must have been or it never would have seen the light of day. About a year ago I did Rhapsody in Blue with a pianist over here called Kevin Cole. He is really renowned as a Gershwin specialist, and was restoring some little tiny four bar cuts here and there in Rhapsody in Blue.


One of the most interesting things was the variety in his performances – it was different from night to night. He would take enormous liberties in the cadenzas etc. I think it’s fantastic. If you’ve listened to the piano rolls of Gershwin playing I can’t imagine he would have a problem with that. This was a guy who started out as a Tin Pan Alley composer, right? They wrote songs, basically showing the kind of harmonic structure and the melody. After that, is it you and a pianist? Is it you and a jazz trio? Is it you and a symphony orchestra? I’m not entirely sure that he would have cared. I think he wrote the tune and the job for the rest of us was to approach it as creatively as we wanted. So I haven’t done that before but I think it would be fantastic, I would love that.


Yeah it really worked. I wasn’t sure if it would, but it was fantastic.


If you were teaching someone who was about to conduct this for the first time, are there any other aspects of this piece you think are important (that we haven’t covered already)?


Well, if you look at p.2 R1 the metronome mark at the beginning of the piece is half note = 69. I tried, I don’t know how many times, after R1 (and after the poco rit) to do that in two, but the poor first violins just did not have enough information. Once I went to doing it in four (which I did eventually), I took it at exactly the same tempo. It was just that they needed more information than I was showing when I was doing it in two. Trying to go ‘Diggadiggadiggadigga Diggadiggadiggadigga Dum’ is harder than ‘Diggadigga Diggadigga Diggadigga Diggadigga Dum’ because then you’ve got four signposts along the way (five counting the downbeat). That kind of thing you just have to be able and willing to do on the fly to be able to sync up.


If I were coaching somebody trying to learn this piece the biggest thing that I would say is (and this goes back to something I said a few moments ago) – if players are thinking more ‘micro’, and conductors are thinking more ‘macro’ then what I think it really behoves you to do is listen to more Gershwin. It’s important to get immersed in the style of whichever composer you’re doing at the time. The first movement, for all intents and purposes, is a Charleston. The second movement is a blues, and the third movement is like an aggressive species of ragtime. This piece was written (and premiered) in 1925 so I would encourage them to go back and listen to Charlestons and blues from the first half of the 1920s, and listen to a bunch of ragtime music – Scott Joplin and all the rest. It’s 2018, we can find all of this online now. That’s what I did as I started to review the piece, getting ready for those performances.


For me, the difficult thing to reconcile was the fifth measure of the piece. It’s Charleston, but when we think of Charleston (well at least over here) we think of a song called Charleston. It really moves. Yet the metronome mark of this music is 69, which is not fast. You have to try and figure out why he’s writing it so slow, and in two instead of four. The best way to figure out why and how that came to be is to go listen to recordings from that time. I wish we could go and listen to ländlers as they were performed in Mahler’s time, or waltzes as they were performed in Strauss’s time. Unfortunately we can’t do that – we do the best we can, inheriting the traditions from the Viennese. But we have recordings of these guys playing these styles at that time, so that technical part of this piece is not really so challenging. It’s more about finding the corners, making sure you know who’s the boss at any given time.


Because our job is more macro, it’s really important that I can adjust the balance in specific places etc. But what’s more important, to me anyway, is that the orchestra has a unified approach to how we’re stylistically going to execute the piece. So in bar 5 (when the Charleston rhythm shows up for the first time) it starts with no accent, beginning mezzoforte, making a quick diminuendo to piano, a quick crescendo to forte and there’s an accent on the second syncopated note. That’s a very fussy marking. I don’t blame him for it at all – I would rather composers be a little too fussy than not fussy enough, because at least you don’t have so much to question. But if you tried to get the orchestra to do exactly what I just said, I mean, you’d drive everybody crazy! It’s better to say ‘everybody just relax a little bit on the first note and then really kind of pop that second note, in a sort of sexy Roaring Twenties way’. Now you’re talking about the gesture, about what it’s supposed to feel like. That to me is way more important than ‘we need a little bit less trombone here, we need a little bit more oboe here’. All of that’s important, of course it is, but if the orchestra doesn’t have a unanimity of approach, then none of the little corrections that you’re giving in these isolated bars are going to mean anything.


Yes. You mentioned bringing out the different colours earlier e.g. the Charleston, blues or even the klezmer moment (at R7 in the second movement). Do you discuss any of that with the orchestra?


Not necessarily, it really depends. When something is very clearly a homage to the Charleston I will tell the orchestra. But it’s very interesting, I don’t ever say ‘you know, bar 5 is a Charleston’. As soon as you say that the orchestra’s thinking ‘no kidding, genius, thanks so much. What else do you have to offer us?!’ There will be a number of people in the orchestra that perhaps don’t know it, haven’t done their homework, or just forgotten that this is a Charleston. So the way that I then present it is ‘you know guys, when we get to the fifth measure, that first statement of the Charleston rhythm? What we really need to do with the rhythm is …’.


Absolutely. It’s all about tailoring the communication.


Right. It’s just being suggestive and allowing that information to get to the players rather than saying ‘this is a Charleston, this is a blues, this is a ragtime. Play it like that’.


Ah, so you kind of ‘drop’ it in. Yes I’ve seen other conductors do that as well, otherwise I think the orchestra can feel a bit patronised.


This piece was also nearly called the New York Concerto. Do you drop in any references to New York imagery?[7]


Well, I will do that for myself, but I’m always very cautious about doing that with an orchestra. The biggest eye roll that a conductor will ever get on the podium is as soon as they say the word ‘sunrise’. They don’t want – nobody – oh my God – ‘oh good a sunrise, someone’s finally composing a sunrise’! They don’t really need that level of poetry. But I do remember saying to the orchestra, particularly in the finale (that kind of agitato ragtime) something to the effect of ‘the original title of this piece was going to be New York Concerto so if we could just have that kind of hustle and bustle of city life, that kind of ramps up the excitement’.


We know what its like to go to New York City or to London in 2018 but we’ve all grown up with it. If you think about being in the Roaring Twenties in New York City for the first time (obviously I didn’t say all of this to the orchestra) think about how overwhelming that would be. For someone that had been born in the Nineteenth Century and then is all of a sudden in New York, it’s going to sound a little chaotic – like the end is nigh. I think that’s the intention. Again, I’ll guide them like that while also trying to provide them with a little bit of information that some of them maybe didn’t have. Hopefully that steers them in the right direction.


What I’m always trying to do is to make sure that we’re getting the composer’s story across to the audience. By saying ‘Charleston’ it just helps the orchestra dance a little bit more. By saying New York Concerto, just in passing, it helps to get some of that chaotic hustle and bustle across to the audience. You know, if you talk about how you’re always striving for that blue note in the second movement but you never quite get there, that actually says so much about the entire aesthetic of the second movement. If you just say ‘this is a blues’ they’ll think ‘what does that even mean? Do you want us to just slide and schmear around from pitch to pitch, is that what you’re looking for?’ So it’s that art of informing them as you’re subtly suggesting…


Well, I hope it’s subtle. Maybe its not!


(laughs) I spoke to Leonard Slatkin about the ‘American sound’ that we often associate with Copland, Bernstein, Gershwin etc. But what he explained was that he doesn’t think the American sound is comparable to the way we approach something like the bel canto style, or the first Viennese School. American music was not initially performed by Americans, and there are so many more influences. Do you agree with that?


I do, I do. One hundred per cent. It’s not that Mozart and Haydn don’t have their individual traits and attributes that make them unique, but I think that that first Viennese school (Haydn, Mozart and early Beethoven) are much more similar than the three big composers you just named from the Twentieth Century in America (Gershwin, Bernstein and Copland). Certainly, there are things that they may share in common, but not the language that they use and the way that they use the orchestra. Bernstein could not have written Appalachian Spring. And Aaron Copland most certainly could not have written West Side Story. I mean, it would all be in A major! I love Copland, don’t get me wrong, but Lenny was a very different composer than was Aaron, than was George Gershwin. So you have to learn to speak their language.


Yes and the Ukrainian/Lithuanian/Russian Jewish influence is there with those three, you’d need to be listening to Russian music as well if you’re thinking about their style/influences.


Yes. It’s that quintessential American phrase, ‘we’re a melting pot’. We take stuff from anywhere and everywhere – you can be as jazzy as you want, as classically European as you want, you can use as much so-called ‘world music’ as you want. Part of the reason I think that you get so much more homogeneity in the first Viennese school than you do in the Twentieth Century in America is there was so much more music by then and it was so much more readily accessible. We had phonograph, all of this. It’s natural that the language expands, but it expanded for everyone in such a different way, which is what I think makes it so interesting. It’s why I love American orchestral music as much as I do, because it doesn’t actually mean anything – other than it was composed for orchestra in America. You could articulate for somebody what Mozart, Haydn and early Beethoven sound like because they sound relatively similar. But I don’t think that’s just in America, I think that’s now true all over…




Greenberg, R. George Gershwin (Phaidon Press, London 1998)


Leon, R. Gershwin (Haus Publishing Ltd., London 2004)


Pollack, H. George Gershwin: His Life and Work (University of California Press, 2006)


[1] It was rumoured that Gershwin had some assistance with the orchestration of this piece from William Daly, a conductor who did the test run at the Globe Theatre. Allan Langley (a freelance violist) claimed that Daly was one of Gershwin’s ghostwriters, and thought that Daly knew the American in Paris score better than Gershwin did (Greenberg 1998:129). Similar claims were made about Damrosch, but Leon writes that ‘the most Damrosch did was to “thicken” the orchestral sound with a little more instrumentation than was in the original draft’ (Leon 2004:65).

[2] In the first published version there was a sizable cut in the first movement (starting ten bars before R8), and from the second movement (at R8). Another cut at R28 in the first movement results in the music jolting to a Gb major triad rather than Db major as originally planned (Pollack 2006:347).

[3] For Gavin Sutherland’s interview on Tchaikovsky’s Swan Lake, click here.

[4] The critics were also rather split in their reaction to this piece ‘into those who saw things beneficial to American art music and those eager to jump on the ‘crudites’ and ‘triteness’ in his scores’. Greenberg felt that ‘the ideal balance’ was achieved in this piece, ‘where the all-American Gershwin grabbed the classical concerto by the throat and poured tune after tune into his metamorphosis of a respected, European art-form’ (Greenberg 1998:163). It was also noted that the audience clapped more at the start of the piece than at the end (Pollack 2006:357).

[5] ‘Gershwin’s lack of academic training – and his yearning to correct it – does not trouble his listeners. It was his own problem, and to a chronic degree. It is a pity the psychoanalysts he came to rely on towards the end of his life could not reinforce something he felt deep-down anyway: that the very process of learning the academic rules of counterpoint to perfection would only go against his innate musical personality and spoil, rather than enhance, the products of his inspiration’ (Greenberg 1998:164).

[6] (accessed 1st October 2018)

[7] The visual element of the piece has attracted a number of choreographers, including Gertrude Kurath (1930), Billy Watson (1981) and Jerome Robbins (1982). Robbins named his work Gershwin Concerto and drew on the New York imagery with one of the themes being ‘urban loneliness’ (Pollack 2006:357).