Piano Sonata Number 21 in C Major, Opus 53 (L'Aurore) (Beethoven)

Saturday, June 18, 2016
Levine School of Music - Strathmore Campus
Bethesda, MD
Difficulty score: 6/9

So I gave my first full-length piano recital in eight years this past June (I had wanted to do it sooner, but trying to finish up graduate school, moving to DC and settling in, and placing a greater emphasis on running these past few years prevented me from doing so).  I realized I had neglected to bore everyone about how during this recital, I finally got to perform this piece that I had wanted to learn and perform since I was twelve years old.  I figured now was a good time to do so, while I'm planning another full-length piano recital for June 2017.

This piece, also known as the "Waldstein" or "L'Aurore" Sonata (I prefer the latter as it is more poetic, even though the "L'Aurore" nickname is not frequently used in the United States), is one of Beethoven's most substantial, popular, and most historically significant piano sonatas.  Beethoven had  composed and published 32 piano sonatas in his lifetime.  This one, composed in 1804, is Number 21 in that cycle and an archetypal example of Beethoven's "middle period" compositions, in which Beethoven started implementing changes in his writing style and amplifying his works to much larger scales.

In the Classical Era, a piano sonata was a large-scale work for piano solo, typically consisting of three or four movements: a fast-paced opening, a contrasting slow movement, (optionally) a dance movement in the form of a minuet, and a moderately to fast-paced finale.  Superficially, the basic structure of this piece adheres to that format: a fast-paced opening movement (0:00 to 11:21 in the video above), followed by an intervening slow movement (11:22 to 14:50) that serves as an introduction to the leisurely rondo closing out the entire sonata (14:51 to 25:20).  But in many ways, this piano sonata was quite different from anything anyone else had ever composed before.  The chord progressions and modulations in the first movement, although they might not seem particularly strange to our ears in 2017, were very novel in the early nineteenth century.  The long pedals in the third movement (the pedal is held down sometimes for as many as ten or more measures), as well as virtuosic figurations such as the lengthy trills (24:25 to 24:59) and the octave glissandos (i.e. gliding the hand across the keyboard; 24:16 to 24:23) creates sounds and pianistic effects that were quite new in 1804.

I first came across this piano sonata some time in 1994 after I borrowed my piano teacher's copy of the complete Beethoven Piano Sonatas so that I could photocopy the "Moonlight" Sonata and begin learning the notes.  I flipped through the entire book to take a look at the other piano sonatas, and this one caught my eye.  The reason was pretty dumb; I was simply fascinated by its page count.  At nearly fifty pages, this piano sonata was much longer than any piece I had encountered in my then limited exposure to the repertoire.  In terms of page count among Beethoven's piano sonata, the "L'Aurore" is also second only to the twenty-ninth in the cycle, the "Hammerklavier."

However, I could not sift through the score of the "L'Aurore."  Editions of the complete Beethoven piano sonatas typically need to be split into two volumes simply because the total page count often reaches seven hundred pages or more, which would result in a book that is too bulky for use at the piano.  The "Moonlight" is often in the first volume whereas the "L'Aurore" is typically in the second.   Shortly afterwards, I borrowed a recording of the "L'Aurore" from the local library, wondering what a fifty-page piece would sound like, and how long it would be in terms of playing time.  Back then, I did not have the patience to listen to all twenty-five minutes of the sonata, but I did listen to the first movement.  I listened to it over and over again, captivated by the soft repeated chords in the opening, the transformation of these repeated chords into a tremolo during its restatement a few measures later, the chorale-like secondary theme, and the rapid broken-chord and scalar passages.  I even tried to learn as much of the first movement as I could by ear.

Not long afterwards, my father purchased a recording of this piano sonata for me.  I started listening to the second and third movements.  I needed a while to warm up to the second movement, a very slow and harmonically unsettling arioso serving as an introduction to the third and final movement, a spacious rondo (a piece of the form A-B-A-C-A-B-A, namely where a refrain is repeated amidst contrasting episodes) that enthralled me upon my first listen as much as the first movement.  I then asked my piano teacher if I could borrow her copy of the Beethoven piano sonatas again; but this time, I requested the second volume so I could tinker with the "L'Aurore".  Understandably, she seemed a little reluctant; after all, this piano sonata is one of the more technically and musically difficult ones of the thirty-two.  She gave in, and I was able to work on learning the notes for a little while before she took the score away, deciding that this piece was indeed too difficult for me.

Three years later, I ordered a copy of the score of the "L'Aurore," attempting once more to learn the notes.  I got much further than I did last time, but since my technique was nowhere near the level needed to successfully execute this piece, I developed many bad habits.  I switched teachers that year, and after my new one gave me a bracing reality check about my current technical and musical abilities, I listened to her, put the "L'Aurore" back on the shelf, and focused my attention on less ambitious pieces more appropriate for my level at the time.  I finished learning all of the notes to the first movement when I was seventeen, and finished learning the notes to the second and third movements five years later right after my graduation from the University of Rochester, after my teacher there suggested I start working on this piano sonata.  I had planned to perform this piece in a piano recital I had intended to give in 2009 that I needed to cancel as I needed to focus on finishing my dissertation and secure a job for after my thesis defense.

Initially, I had intended to perform another Beethoven piano sonata during this recital, but after some consideration, I decided now was a good time to perform "L'Aurore"; I had brought this up to performance standards before, after all, and the notes came back to me surprisingly quickly; after less than two weeks of work on the "L'Aurore," I was able to play the first movement at tempo.  Plus, I felt this piece would be more friendly toward an audience that was able to appreciate piano music but did not know much about it.  But I'm happy to say that I can cross performing this piece off my bucket list.

Considerations for Learning and Performing this Piece

• I gave this piece an overall difficulty rating of 6 out of 9.  This rating is primarily because of the scale of this sonata and the technical figurations; this is one of the virtuosic middle-period Beethoven sonatas, even though it is not the most difficult piece Beethoven wrote.
• Practice slowly, of course, and figure out how to convincingly integrate the many virtuosic passages into one cohesive whole.
• Focus on clarity of the notes and keeping a constant tempo throughout each of the movements; this is still a Classical Era piece, after all.
• In the last movement, the opening section (14:51 to 16:10) recurs several times essentially verbatim.  Figure out a different way to execute this theme each time it appears to change things up and to prevent the piece from getting too repetitive.  During the first occurrence, I like to milk every note to really emphasize the spaciousness of the theme.  During the second occurrence (17:20 to 18:43), I play more flatly.  During the third and last occurrence (21:29 to 22:07), I play the theme more grandiosely to emphasize its glorious return after a lengthy and tempestuous middle section.

Race Post-Mortem: Rock and Roll DC Half-Marathon 2017

Some of the half-marathoners and marathoners from DC Front Runners, freezing at the finish line.

Saturday, March 11, 2017
Washington, DC
Difficulty score: 5/9

It's been a while hasn't it?

Anyway, this past weekend was my first race of the 2017 calendar year and my first race in over three months.  I had intended to run this race as a tune-up for my very first Boston Marathon; not only would this help me get back into the racing mindset after a three-month hiatus, but it would also serve as the basis of a target marathon time (although a conservative one given the difficulty of this course).  I had done this half-marathon back in 2015; based on that, I decided that I should not be surprised if I finished one or two minutes slower than expected.  Thus, before the race, I had settled on 1:26:59 as a baseline goal; 1:25:59 or better was my "if everything goes right" goal.

Training

So far for this training cycle, I had essentially repeated by training plan for the Richmond Marathon this past November, with a basic weekly structure of:

• Monday: five to seven miles easy (sometimes slower than 9:00/mile) to recover from the previous day's long run.
• Tuesday: tempo run days.  These would typically consist of six to eight miles with four to six miles at 10-miler to half-marathon pace.  I ended up covering the tempo portion between 6:15/mile and 6:30/mile most of the time.
• Wednesday: four to five miles easy to recover from the previous day's tempo run.
• Thursday: longer mid-week run between eight and eleven miles, some of them with progressions to target marathon pace (7:00 to 7:05/mile) or faster.
• Friday: OFF
• Saturday: seven to nine miles easy (7:42 to 8:24/mile pace).
• Sunday: long runs of up to 19 miles.  Typically these miles were covered at the same pace as Saturday's runs, but some of them involved progression to target marathon pace.

So far, I have been able to complete most of my training as written; I have run at least 50 miles per week most weeks these past few months, with a maximum of 59 miles.  I probably would be able to get one or two more weeks of mileage in the fifties done before Boston.

The Race

March is apparently the new January this year.  Race day weather was sunny and 26 degrees Fahrenheit, with a 15 mile per hour wind: definitely the coldest conditions under which I have run a half-marathon.  But 26 degrees was much better than 76 degrees, and I certainly preferred these conditions to the steady cold rain of the 2015 race.  Besides, once I started running, I didn't really notice the cold too much.

The course had also changed slightly from the last time I had done it.  At around Mile 2, after the DC Front Runners water stop and the Lincoln Memorial, the previous route involved an out-and-back segment across Memorial Bridge to the rotary right in front of Arlington National Cemetery.  This was eliminated this year and replaced with an out-and-back segment along a brief freeway leading to I-66.  The race would also end right in front of the DC Armory, instead of in a nearby parking lot as was done in 2015; as a result, instead of a downhill finish, we now had a slight and steady uphill one.

The rest of the route was similar to the one from 2015.  We started at 14th Street and Constitution Avenue, going toward the monuments (with a brief dogleg near the half-mile mark) and then turning on 23rd Street toward the Lincoln Memorial.  We then would go down Rock Creek Parkway toward Northwest DC, through Adams Morgan and Columbia Heights, and then along North Capitol Avenue south toward the Capitol for a somewhat hilly finish through Capitol Hill.

Of course, that nasty hill leading out of Rock Creek Park and into Adams Morgan at around Mile 6 was still part of the course.  The good thing about this hill was that it was short.  The bad thing about this hill was that it was quite steep; it was a 140-foot climb over one quarter of a mile, which could really disrupt your race if you don't approach it carefully.  Thus, my plan was to treat this race kind of like a 2 x 6 mile run: I would stick to half-marathon effort for the first six miles before the big hill (and not freak out if my pace wasn't as fast as it was in Shamrock last year), jog up the hill and for the next half-mile for recovery, and then return to half-marathon effort for the last six miles, hopefully closing with an even faster pace.

Map of the new route and the elevation profile (from the race website).

Miles 1 through 6 (6:33, 6:32, 6:32, 6:34, 6:34, 6:27): The first half mile was spent weaving through a lot of people and trying to get onto pace.  My pace slowed when we turned off Constitution, but then when we were about to get back onto Constitution, I took advantage of this small hill to speed up, completing the first mile a little faster than I thought I would: 6:33, when I was expecting to finish it around 6:37 or so.  But what did not make me happy was that I was already running the tangents terribly; even at the Mile 1 marker, my Garmin indicated that I had completed the first mile several seconds before I actually crossed the mile marker.

Mile 2 was relatively flat, with some downhill portions along Rock Creek Parkway, so I spent it trying to keep my pace under control.  I waved to the DC Front Runners as I passed their water stop.  Miles 3 and 4 featured that new segment along the freeway.  I found this slightly more difficult than the out-and-back segment along Memorial Bridge to Arlington National Cemetery, but nothing too outrageous; just a bit more rolling hills.  Miles 5 and 6 were the same as the old course; just down Rock Creek Parkway toward that hill leading up to Adams Morgan.  During these miles, I tried to keep an honest pace, yet tried to maintain an even effort on both uphills and downhills.  I tried not to change up the uphills to conserve energy for the hill coming up.  I was pleased to find that my splits for the first five miles were pretty consistently under 6:35.  Since Mile 6 was mostly downhill, I was able to cover that mile even faster.

Mile 7 (7:16): As planned, I turned race mode off for the hill.  I slowed down, pulling myself up the hill and giving high-fives to all the people standing on the side of the course.  I just focused on getting to the top and not worrying about how quickly I was doing it, even though I really wanted to be done with it as soon as possible.  But it turned out that nobody passed me on this hill (but then again, I was also a few hundred feet ahead of the next runner).  And my pace on this hill was about an 8:20/mile; not nearly as depressing as I thought it might be.

After I finished that hill, I ran the next half-mile easy also in order to recover from that effort.  This was something that I decided to try this year.  Last time, I tried to get right back on pace immediately after climbing the hill, which was most likely a mistake; my legs felt drained for the rest of that race.  After climbing the second slight hill on Calvert Street leading to the center of Adams Morgan, I started to pick up the pace again.  Going into the race, I accepted that I would finish this mile quite a bit slower; I decided I would have been happy if I finished this mile in around 7:15.

Miles 8 through 11 (6:32, 6:24, 6:21, 6:16): I was able to get back on track pretty quickly, even in spite of another substantial hill near Georgia Avenue.  But after I crested that hill, I knew I had a fast part coming up on which I could make up a lot of time; Miles 9 and 10 had some very long downhill segments, and I took advantage of this to push the pace a little.  I tried not to overdo it since I remember Mile 12 being another difficult segment.

When I crossed the 10-mile marker, my watch read 1:06:21.  Unfortunately, my watch also indicated I had covered 10.11 miles.  I quickly did some math about how fast I needed to run the remaining 5K.  I was definitely going to be under 1:26:59.  If I kept my pace under 6:27/mile for the remainder of the race, I would finish in 1:26:21, and I would be fine with that.  Anything beyond that was a bonus, and finishing in 1:25:59 or faster was still not out of the realm of possibility.  Mile 11 was still relatively flat, so I pushed the pace a little more; I was able to cover it in 6:16 with a little energy left to get me through the next mile.

At this point in the race, I was beginning to feel a little warm; either shorts with the long-sleeve black T-shirt or a singlet, arm-warmers, and running tights may have been better options today, even in spite of the cold.  I could have also removed my hat, but I didn't feel like carrying it in my hands as I was running.  Plus, I was having a bad hair day and didn't want that immortalized in my race photos.

Me at around Mile 7 in Columbia Heights.  This is what I was wearing this race, and I thought it may have been a little much.

Miles 12 to the end (6:27, 6:08, 1:30 for the last 0.26 miles): The first half of Mile 12 is generally flat whereas the second half is a steady uphill, though not as stringent as the ones seen in Adams Morgan and Columbia Heights earlier in the race.  However, the fact that this appears so late in the race makes it suck quite a bit.  I held my pace steady, trying to conserve just a bit of energy; I decided I would not freak out if my pace for this mile was quite a bit slower.  I did have the mostly downhill Mile 13 to make up some time.

I covered Mile 12 in 6:27, but I still managed to catch and drop several people on the uphill.  Knowing that much of Mile 13 was downhill (but not realizing that due to the changes to the course, the last 0.1 miles were slightly uphill), I pushed the pace a little.  I no longer paid much attention to my pace; I glanced at my watch every now and then to see how much time I had left if I wanted to break 1:26:00.

Thanks to the downhill, I managed a 6:08 mile for Mile 13, but struggled a little when I came to the last uphill before the turn to the finish.  Jeremy was on the side of the course cheering on the half-marathoners and marathoners (at this point the two courses merge together for the finish), which was a very nice surprise and gave me the last bit of energy to carry me through the finish.  As soon as I rounded the corner, I gave it everything I had, hoping that I would cross the finish line before my watch read 1:26:00.

According to my watch, I finished in 1:26:04.  But while waiting for my friends Alfie and Ryan to finish the marathon, I discovered that my time was actually 1:26:02 (6:34/mile).  Apparently, I started my watch too early or stopped it a second late.  Even though I would have really liked to have finished three seconds faster, I think this is what I had in me today.  Plus, it's 71.44 points toward my DC Front Runners Race Circuit score (age grade score times 1.05 for a half-marathon in accordance with DC Front Runners Race Circuit rules), not to mention more than one minute faster than my time for this race in 2015.

And I really did not do a good job of running the tangents today.  I ended up covering 13.26 miles total according to my Garmin.  I did think about whether I could have done more like 1:25:30 had I covered 13.16 miles rather than 13.26.  But then again, this is kind of a twisty, curvy route, unlike the Richmond Marathon or Shamrock Half-Marathon routes.  This could be one of those routes in which it is difficult to run tangents well.  I don't think weaving in and out of people at the beginning helped much either.

5K, 10K, and 10-mile splits from the runner tracker that the race provides.  Looks like the hill didn't slow me down too much.

Next Steps

I have the Cherry Blossom 10-Miler in three weeks.  I'm using it as another tune-up race before Boston and as another opportunity to score some points toward the Race Circuit.  It is close to Boston (only fifteen days before), but plenty of people in my running club have done both Cherry Blossom and then Boston with little ill effect.

Now that Boston is less than six weeks away and that I have a half-marathon tune-up race under my belt for this training cycle, I suppose now is a good time to start thinking about target times.  1:26:02 times 2.15 is 3:04:58 (I used the multiplier of 2.15 given my mileage this cycle).  I would be satisfied with that, since it would be a BQ-5:00; I would love to be able to register during the first week of registration for Boston 2018.  Perhaps this might even be a little conservative given the difficulty of the Rock and Roll DC Half-Marathon course and the conditions on race day.

Of course, these predictions are contingent on the weather in Massachusetts on April 17, 2017.

Other Thoughts

• This course can really eat you alive if you are not smart about race execution.  I think the main thing about this race is patience; I would recommend trying to keep the early miles under control and focusing on conserving energy, not charging up that hill, and then using some of the later miles to make up the lost time from the hill and the segment through Adams Morgan.
• This year, they started the marathoners an hour and a half earlier at 7:00 am, possibly to alleviate some of the congestion at the beginning.  One advantage of this was that some of my friends doing the marathon would finish about half an hour to an hour after me; I got to meet them at the finish without waiting in the cold for such a long time.
• If the cold did affect me, I don't think it was by much.  Waiting for the start did not feel too terrible and even though it took a little longer to warm up, once I hit Mile 5, I was fine; as I mentioned, I might have even been a little overdressed.  I did really feel the cold after the race, though.
• Recovery from this race is taking a little longer than usual; as of Tuesday evening, my legs still felt tired and sore.  Usually, I'm feeling better after a half-marathon at that point.