Figura 1 Updated Master inventory of the ARS Piano Rolls. Document provided by the Archive of Recorded Sound
Esther Burgos-Bordonau
The Spanish Piano Roll collection of the ARS was the subject of our interest from the very first moment we came across it while browsing their website. As mentioned before, the ARS is considered one of the largest and most prestigious sound archives within the USA and collects thousands of musical examples in all types of media. Our latest research interests are focused on Piano Roll collections and one of our first contributions was the paper presented at the Spanish National Congress of Musicology, held in Madrid in 20161. In those days we were especially interested in Zarzuela rolls, one of our most characteristic genres of Spanish musical styles. Since then, our research on Piano Rolls has continued, being this specific collection a turning point in our research trajectory2. Before applying to Berkeley, we had already checked Stanford University and its libraries and we must thank the Complutense University of Madrid, which allowed us to research Berkeley University (and all the Bay area, Stanford included).
In the very beginning, the Archive of Recorded Sound had some interesting rolls, which had come from donations and acquisitions, and all these items are described under the ARS0141 call number. But the first collection that significantly increased this Archive of Piano Rolls was the Denis Condon Collection of reproducing pianos and rolls. This outstanding collection was probably the main reason to start with the Piano and Organ Roll Project that Stanford University launched in 2015, after this collection had arrived at their Archive from Australia3. However, a few professors of piano music from Stanford University were also very interested in performance practices of the late nineteenth and early twentieth centuries4. This Piano Player Project promotes study and research into all the aspects related to performance in the past. They are convinced that this exciting project will bring together researchers, musicians and other people interested not only in these old melodies but in the way in which they were played. It’s a significant source for the early recordings as most of these were made around 1900 and 1905.
Some Symposia were held before this project started, which concluded that sound recordings and the study of historical performance were an exciting way to learn much more about how pianists played in the nineteenth century, as there are many recordings which were done by the composers or other famous pianists5. These are known as ‘artist rolls’, and they can be the first and foremost source to accurately discover how George Gershwin, Igor Stravinsky, Claude Debussy or Enrique Granados, played the piano6. These are considered their early and reliable recordings as cylinders could not last more than 4 or 5 minutes and the quality of the recording is not comparable. Not even shellac discs, which were also shorter in time, and when recordings were done, the pianist had to stop so that the disc could be turned over (in case the disc could be recorded on both sides) and then continue playing. Therefore, Piano Rolls soon became the best way to record an entire piece or at least an excerpt of one. Since that first event, other international Conferences have been held. The first “Global Piano Roll Meeting” took place in Lepzing in 2018. After such an interesting meeting reuniting all the specialists, musicians, music librarians and enthusiasts, the aim of the Conference was to be held in a biannual basis7.
So the vast collection of Denis Condon (1933-2012), an Australian musician, composer, professor and educator who fell in love with these kinds of pianos since he was fifteen, till his death in 2012, because his father bought one of them for their home, was the main reason that moved Stanford University to start with this ambitious and vital Piano Roll Project, later known as SUPRA (Stanford University Piano Roll Archive).
Denis Condon’s desire was to place the whole collection, instruments included (and some other materials such as tools and parts for repairing the devices, historic advertising materials, correspondence, etc.) in an institution that could make these rolls work, which would be useful for musicians, scholars and people interested in this kind of unique music material.
Stanford committed to going on with this work not only accepting this enormous collection, which was purchased by an anonymous donation in March 2014, but cataloguing, describing, preserving and digitizing as much as possible to offer it to the music community and researchers. It arrived from Australia to Stanford in August 2014. There are some excellent videos in Youtube showing all this incredible removal8.
When the Denis Condon collection was presented to the audience in February 2015, Mr MacBride mentioned in the MLA Conference held in Denver (CO) the main challenges for Stanford accepting this vast collection9. These were to make possible playing all those old rolls (most of them in a perfect state of conservation), the maintenance of the instruments (usually abandoned by the institutions) and to preserve them in the best condition.
The last but not the least exciting chapter of this whole project is the digitization of all the rolls and the conversion of those images in MIDI files so that the researcher will have the opportunity to see and listen to those old melodies.
From the very first moment when Stanford received this collection, they have been hard at work, cataloguing and categorizing. When the collection was given in February 2015, and during all these years, a lot of work has been done on this collection. We have recently checked that some ‘artist rolls’ have been digitized and are available through Stanfords’ webpage. A beautiful example is the “Children’s Corner” from Claude Debussy, played by the composer. This record can be full screen seen and also downloaded the digitized Piano Roll as well as the MP4 sound10. The result is incredible and the sound very beautiful.
The Denis Condon collection (ARS0163) has 7450 rolls. This collection could have reached 7540 rolls and more than ten different instruments, but not all of them came to Stanford. These consist of two grand pianos (Marshall & Wendell Ampico A and Chickering Ampico B); four upright pianos (two Steinway Red Welte pianos, one without keys –because of the import problem of the ivory– and a Rönisch Animatic Piano Liszt Hupfeld Triphonola, and a Behning Art Echo). Also four cabinet players or Vorsetzer (Red Welte, Green Welte, modified to play Welte licensee rolls, Ampico and Duo-Art) too. But, apart from this extraordinary collection, the rest of the Piano Rolls that Stanford keeps at present are the following collections:
Around 17.963 rolls have been inventoried which makes the ARS of Stanford one the most important sound archives within the USA11. The next image shows the the master inventory of the Piano Roll collections kept in the Archive nowdays.
Figura 1 Updated Master inventory of the ARS Piano Rolls. Document provided by the Archive of Recorded Sound
During the autumm of 2018, we were granted by our University, Complutense University of Madrid, to go to the United States and start a new research project. We had already started studying different Piano Roll collections in our country but this new ‘American’ collection was to be a new challenge in our research. Before our arrival at Stanford, we had already checked on the webpage that the ARS was collecting this unique collection (The García-Sampedro Piano Roll collection) that quickly caught our attention. Not only because of the Spanish or Hispanic name (at the beginning we could not know the nationality of the donor) but also the repertoire represented through the rolls.
Before starting the task with the collection, I was given a general inventory with a brief description of all the different collections that the archive had, similar to the one shown before, but this time focused on the Spanish collection.
The collection had to be asked in advance, as most of the materials are kept off campus. However, the first day of work at the Archive the whole collection was prepared and waiting for us. Three acid-free boxes contained the entire collection of 54 items.
Figura 2 The María Jesús Casado García-Sampedro Piano Roll collection. Property of the ARS. Stanford Library
We were kindly helped by the members of the staff who not only gave us all kind of details of this exclusive collection but also brought for us a particular machine explicitly made for opening, unrolling and handling the rolls. It was simply great. Never before had we seen something so simple and useful at the same time.
Figura 3 Table for unrolling the Piano Rolls. Property of the ARS. Stanford Library
Figura 4 Roll inserted before unrolling. Property of the ARS. Stanford Library
Figura 5 Unrolled item ready for study. Property of the ARS. Stanford Library
With this incredible ‘table’, seeing and studying the different marks and notes printed in the roll was an effortless task. It allowed us to unroll the whole document and observe each one of its features. Both handles had to be moved at the same time and with a certain rhythm to unroll it, and once you had seen it, you could rewind it entirely in exactly the same way.
Because the rolls were in good condition and we must say that most of them were, we were able to look at almost the 54 items. However, if the flanges or hubs were broken, the rolls could not be unrolled or played in a mechanical piano. These pieces of the roll, or sometimes the missing hook, were essential for a proper playing of the roll. Therefore, the work of curators and other specialists in fixing and arranging these types of items is crucial for the preservation of these materials. Fortunately, the Spanish collection was mostly in excellent condition.
We had some previous experience in researching Spanish Piano Rolls and Spanish collections. Since 2014 we were involved in some exciting projects launched by the Spanish National Library (BNE). After a couple of years studying their Piano Roll collection – which in those days was around 6000 items – we were able to present a paper in the Musicological National Congress (Congreso de Musicología de la SEdEM) held in Madrid in 201612. After this first approach, we had the great opportunity of meeting the heirs of a well-known family of writers and poets of our Spanish Zarzuela and other dramatic works, the Fernández-Shaw family. This casual and fortuitous encounter gave us another possibility of learning and studying more about Piano Rolls and the standard repertoire which was represented, at least in Spain, during the thirty-first years of the twentieth century. Our contribution to the knowledge in this field was published in different journals, congress proceedings, an exhibitions’ catalogue and a final work with the complete description of the entire collection.
As a result of this experience, when we decided to work with the Casado García-Sampedro Piano Roll collection, we were at least prepared to deal with titles, artists and styles of music that we had previously seen in other Spanish collections. Our first hypothesis was that we were in front of a Latin American collection. After our first interview with the Head of the Music Library, we knew that it had belonged to a Spanish lady until 2015 when she decided to donate it to Stanford University.
Once the Latin American path was discarded, we had to find the meaning of this Spanish collection and discover which were the items that made it ‘rare’ from the others kept in the ARS. Soon we could confirm that it was different to their collections not only because of the composers, many of them Spanish musicians, but mainly because of the different genres of music so typical in Spain in those days. Another important aspect were the publishers, mostly all Spanish ones, with typical metronomical rolls.
After seeing the brief but almost complete description offered by the Master inventory of Piano Rolls given to us by the ARS, in terms of information, we decided to fill, as much as possible, all the gaps we could see from our point of view. So we decided to complete the names of the authors, and titles of the works and subtitles when possible; dates of birth and/or death of the composers; publishers, some were unknown, but we were finally able to discover them and the most important information for us, which was missing in this list, the musical style/genre. At this point, we could confirm that an essential number of items were Zarzuelas (a typical Spanish genre) or other musical styles very popular in the twenties and thirties in Spain and throughout Europe13. This was music for entertainment such as Waltzes, Polkas, One-steps, Fox-Trots, Charlestons and other dancing styles14. We also found different South American styles like Tangos or Pericones. Apart from these, there were also items with classical repertoire: Variations, Sonatas, Hymns, Operas and Piano Music.
Figura 6 Excel document with the Master Inventory of the ARS Piano Rolls. Document provided by the ARS Archive for the research15
Once we realized our first hypothesis was wrong and could confirm that it was a Spanish collection, we set several goals. The first one was to get in touch with the donors. The Head of the Library helped us to write to the donors and to explain to them our work in the ARS and our interest in getting in touch with them. Of course, due to the data protection law and the confidentiality towards its donors, the Head of the Library was the only person who could introduce us to the family. We then had to wait and see if they were interested in contacting us. If this happened, we would have the opportunity to know not only more about them and their collection but also the reason why these items had ended up at the library.
We were really surprised to find a Spanish collection in Stanford, especially after seeing how the National Library of Spain had made such a great effort in trying to preserve and keep private collections in the last ten years as they are very concerned with Spanish heritage and with all those works that shouldn’t go out of the country. However, as this was a private collection we finally understood the reasons that moved them to donate their rolls to this extraordinary Archive, though we cannot hide the fact that we would have liked to find it in our National Library.
The Head of the Music Library explained to us the vital advertising and marketing campaign they made of this Piano Roll Project, not only in the USA but also around the world and that’s how the Spanish donors knew about this great project and decided to give them their entire collection as a gift. As soon as we found this out, we were amazed as we could not imagine the family was not going to look for somewhere closer, such as Spain, to donate the collection, but at the end they decided to send it to Stanford. Today Stanford University, the donors and also the researchers must congratulate them for this decision as the collection is being well conserved, will be better preserved and, most importantly, it is being studied, catalogued and described and will soon be spread worldwide through the Music Library’s webpage. Later we will explain this more thoroughly.
Once that we knew that the donors were Spaniards, our first intention was to contact them to understand the complete history of the family, the collection and all the materials that we could access. It was also crucial to know where they were living now and if it was realistic to meet them.
It was imperative to know when and how this collection was acquired; who was the principal owner María Jesús Casado? Any other member of her family? Why Piano Rolls? Did they have a pianola or any other mechanical piano at home? If so, where did they buy it? When did it happen? A lot of questions to be answered.
The second and crucial step in our methodology was to look at all the bibliographic sources about music composers that were important within the collection. Some of them were popular but others not so much and there were composers entirely unknown to us. Therefore, looking at reference tools such as dictionaries, encyclopedias and all other music materials (books, articles, brochures, concert handheld programs) it could be an exciting way of researching more about this composer and the style of compositions for which they were well known.
And finally, compiling and assembling all the documents property of the family would be an excellent source to build the whole story of this collection. But this last chapter wouldn’t be possible until step number one (meeting or contacting the donors) could be undertaken.
As we had exactly two months for the initial part of this project, to gather as much information as possible, we decided to follow a timeline to organize the work. We were aware that after leaving the Archive, any details about the materials would be very difficult to check. So we decided to organize the entire time doing different tasks in the following eight weeks: Data acquisition and collection (first two months); Creation of a database to introduce all the information collected; Creation of the social and personal context of this collection. This work would be done back in Spain, after our placement; Completion of the standardized bibliographic catalogue, including all the missing data concerning authors’ names, complete titles, uniform titles, and the notes area, essential for us as this field contains the most relevant information of each roll; Final drafting and sending of material to Stanford ARS; The scanning of the rolls and transcription of these to MIDI files (work done by Stanford Library) and the final assembly and launch of the e-book in Stanford’s Library website16.
With this exhausting but relevant chronogram, we started our work on the ARS from the very first day, and we can confirm that the first two points were made on time. We were able to see the entire collection several times, check all the rolls one by one, unroll them with the special table and take multiple pictures of all of them. We also signed the proper documents for making these images, which should be spread and disseminated only for research purposes.
Taking photographs is always an essential part of our work as it helps us to identify each one of the items and also outline the most exciting features on them. We took photographs not only of the rolls themselves but also of the containers: boxes that were not always in perfect condition. Some of the lids were missing and some parts of the boxes broken. We can confirm that the rolls were better preserved than the boxes. It was precisely at this point when the work started to get a little bit complicated. We could check that some of the rolls were not kept in the correct boxes (informing the staff members of the Archive of this) and the most curious part of all came when we found some ‘lost’ labels or ‘stamps’ in different boxes where they were supposed to be. We also informed the librarians of all of these incidents, which gave us complete freedom to work with the collection and to organize it as best as possible. We finally took 368 pictures in jpg format of the entire collection. These included all the rolls (54 of them), their main features, labels, leaders, boxes, stamps and seals and other significant aspects involved in the punched paper.
Figura 7 Eighteen Rolls, of Box nº 2, displayed on the table. Reading room of the ARS
Figura 8-9 “La Marsellesa” by Rouget de L’Isle and the box with the label of “Aida”, by Verdi
Another essential aspect that was very necessary for our initial part of this detailed work was to give a number to the boxes. Although they were just three, we wished to know and to describe better all the rolls that were kept in box number one, box number two and three. That was one of the columns that we decided to add to the inventory (an Excel page) given to us by the ARS, as the first tool with which we could start working.
We also decided to skip some other columns of this Excel page which had made no sense in this Spanish collection. For example, roll number alternative, original or reprint, or the player. Instead, we created the first column with the box number, and we decided to give a number to each roll (from 1 to 54). This factice number was added in pencil to each one of the anti-acid papers that covered the rolls (we were also authorized to write this on the paper) so that we could easily find a roll inside a box and identify it quickly in our initial inventory. After this first task we could know that box number 1, box number 2 and 3 had all 18 rolls each, though the sizes of them were not always the same. That was our next challenge.
Indeed, some boxes were different from others. Some were bigger because the rolls were also thicker. That means that though some were standard rolls of 5 or 6 minutes and eighty-eight notes, others were longer, and that is the reason why they were also thicker, they lasted longer. These thick ones were usually classical repertoire such as Beethoven (roll n. 39, box 3, roll number 1894) and some others. Another example of this full roll was the one titled Très Jolie by Emile Waldteufel (1837-1915), which was not only thicker but was rolled in a wood matrix, entirely different from the rest of the collection and also with different sizes. We, the researcher and piano professor assisting me at that point, worked out that this was a 65 note roll, the only one of this type in the collection. It was unique and also impossible to be played17.
Although we have said from the very beginning of this description that most of the rolls were in perfect condition at least 20 of them had some problems with the flanges or with the hooks, which meant they could not be played. However, we must also add that at no time during our initial days of research, were we thinking of ‘playing’ these rolls. In fact, the librarians spoke to us about this. The collection, due to its singularity, could not be played in any of their multiple piano players.
We were not surprised as it is not very common to have the opportunity to study and listen to the objects of the research but, surprisingly, a piano professor came to visit me with fantastic news. We were going to try to play some of these Spanish rolls in his beautiful Ampico piano player in his office. We were very excited with this possibility (as no one before had even thought about this chance), and for this first session, we just chose 4 or 5 rolls, those in the best state of conservation. It was a complete success. Here is when we noticed the singularity of some of the rolls.
Thanks to this test we could see how the roll of 65 notes did not fit, in any way, in the Ampico piano player. But many others could, and the unique experience of listening to these old melodies was delightful. From that moment we both decided that we were going to make some ‘live’ recordings, always with the permission of the Head of the Library who was also very happy with this possibility.
Figura 10 Mason & Hamlin. Ampico piano player. Property of Stanford University, Music Department.
So, after this fantastic discovery, which enabled us to play some of the rolls of the Spanish collection, we managed to make twenty live recordings. These are now kept in Stanford Music Library, together with the pictures.
After our two months of intensive work in the ARS and much more work done home once we returned, the final descriptive catalogue was finished and completed. We had the possibility of checking some of these rolls with others in Spanish libraries, especially the National Library of Spain, that is the primary holder of these types of musical documents. This comparison helped us a lot in our detailed work and also permitted us, not always but at least in twelve cases, to listen to the recordings, as our National Library has also digitized around 3.644 items of their whole collection, which today is around 7.000 rolls18.
Having all those pictures with interesting information of each one of the rolls allows researchers to go further with their research into pianola repertoire, publishers and workshops and where these instruments were built and sold. We can learn more about cafes, concert halls and other places of recreation where the pianola became the center of attention. Particularly in Spain, its diffusion began in good measure thanks to a pianola located in a famous spa in La Garriga (Barcelona - Spain), but this is another crucial and long chapter of another different research.
To summarize, the Maria Jesús Casado García-Sampedro Piano Roll collection is an exciting and ‘rare’ collection amongst all the different types of rolls that the Archive of Recorded Sound of Stanford University keeps. It is the only one with a Spanish repertoire of mostly Spanish composers. It also has other dancing styles typical of Spain and Latin America, such as Pasodobles, Tangos, Javas, Lagarteranas, Fox-Trots, Charlestones, Sardanas, Peteneras or Couplets. It also has some classical repertoire, but most of it corresponds with Zarzuela (typical Spanish genre) and Tangos (typical Argentinian style)19.
The challenge of studying this collection opened the possibility of reexamining these compositions and excerpts of Zarzuela again, or other major music works. It has also permitted us to listen to these not so old melodies and retrieve those names and styles for future researches, not only with a complete bibliographic catalogue but also with the digitized images and MIDI files of those rolls complete with the whole story of the collection, the donor and all the details that finally brought it to Stanford Archive of Recorded Sound.
This research20 has been a small step forward in the knowledge of the music composed for pianola rolls and in the predominant tastes and styles of the time, allowing us to learn more and more about the authors, artists and performers involved in this type of musical trade, as well as the owners of these unique private collections. All of them allow us to draw a very singular musical panorama, characteristic of the first thirty years of the 20th century.
Articolo proposto il 2 ottobre 2023 e accettato il 4 novembre 2023.
ESTHER BURGOS-BORDONAU, Universidad Complutense de Madrid, e-mail: eburgos@ucm.es.
Last website consultation: September 27th, 2023.