edit in Google

Examining the Carnegie Hall Booking Ledgers (1955-2007)

Between 1955 to 2007, Carnegie Hall scheduled performances in a set of 52 paper ledgers. Though the Hall started handling scheduling with software in the late 1990s, these ledgers were used concurrently as backups before being discontinued in 2007.[1] These ledgers contain many layers of information: we can see the variety of performers, genres, and non-musical events that took place at Carnegie Hall throughout the decades. Through these ledgers, we can explore control and coordination in gathering information in a slightly decentralized organization like Carnegie Hall, and how paper can act as a way to reconsolidate information. These ledgers also provide a lens to examine the efficacy of paper as a means of gathering and storing information, especially as internal customs emerge and modifications are made.

Figure 1.1

The ledger consists of templated forms (see Figure 1.1). This ledger shows performer names, what date and time they performed, which hall they performed in, when the event was scheduled, and the agreed upon rate between the performer and Carnegie Hall. Each page is separated into sections based on the days of the week with an adjacent line to fill in the exact date. Under each day, there are three columns, one for each hall within Carnegie: the Main Hall, the Recital Hall, and the Chapter Hall. These halls were later renamed Stern Auditorium, Weill Recital Hall, and Zankel Hall, respectively (see Figure 1.3). Under these columns, there are divided rows referring to event time to fill in the name of the performer, labelled ‘M’ (morning), ‘A’ (afternoon), and ‘E’ (evening). For the Saturday and Sunday entries, there are additional rows labelled ‘5:30’ and ‘Midnight.’  For each of these rows, there are also two boxes next to them labelled ‘Closed’ and ‘Rate’ under the heading ‘Lease.’ ‘Closed’ likely refers to the date the performance’s scheduling was finalized between the performer and Carnegie Hall, while ‘Rate’ refers to the amount the performers are paying to perform at Carnegie Hall. Other elements of note on these ledgers are: 1) the designation of a red circled ‘L’ on some entries (possibly a designation to only open seats on the lower level), 2) an unlabelled number on some entries which is sometimes circled on ledgers of later years (possibly a designation of how many seats to open to the audience), 3) parts of some entries have been redacted when the ledger was scanned and digitally uploaded, and 4) many pages show edits to the schedule with events being crossed out and replaced with new ones.

        One question that these ledgers raise is how control and coordination can be managed when gathering information -- a question that becomes more complex because the ledger itself is a means of coordination. From just the one page shown in Figure 1.1 (from the first volume created in 1955), there are at least five different handwritings visible, indicating that scheduling these events is a collaborative effort that requires coordination between multiple people working at Carnegie. Given that there are 52 volumes with over 100 pages each that stretch from 1955 to 2007, there would have been dozens of people (if not more) who participated in this information gathering process throughout the years, and many during the same time period as shown by the multitude of handwritings on one page.

        Though not explicitly stated, for an event to be confirmed in Carnegie’s schedule, there presumably must be administration beforehand: review and approval from management, the payment of a deposit, etc. Carnegie hosts its own concert series where it invites artists to perform, but it also operates as a rental hall where people can simply pay to host their event at one of Carnegie’s three venues, and these rentals make up a large portion of Carnegie’s programming. In Carnegie Hall’s own description of the booking ledgers collection, it is also mentioned that there is a “Booking Director” who decided to use the paper ledger alongside scheduling software as a back-up. In addition to dictating how to record the schedule, this booking director also likely has some say in what events are approved and hands down the responsibility of creating entries in the ledger to others. These factors show some degree of top-down organization in creating the ledger.

        While the format and steps to create an entry in the ledger seem somewhat top-down, the ledger itself and how entries are made and edited show much more collaboration across a team. As mentioned earlier, the different handwritings suggest that many people were given access to modify the ledger. These people would have used each other’s entries to check whether events conflicted or certain time slots were open, making it important that everyone responsible for entering events did so in a timely manner. Though this ledger may have been accessible for other employees at Carnegie Hall to check the schedule of events, the ability to add or change information in it likely would have been limited to a select number of people.

        The redactions made when these ledgers were put online also display another layer in the discussion of control and coordination, as this is essentially another layer of information retroactively added during the upload of the ledgers. It leads us to ask who made the decision on what information had to be redacted -- the person doing the redactions, or someone above them who gave an order for certain information to be redacted?

        The richness of the ledger raises another topic of discussion in the efficacy of paper as the medium for recording and storing information and how much it allows for change. Though the printed template provides some structure in terms of how events should be organized by venue and date, the entries themselves show some signs of unstructuredness. For instance, not every entry is diligent in noting the close date and rate. In addition, the nature of the template itself also is somewhat vague; the range of times that morning, afternoon, and evening could be referring to is quite vast. This led to the emergence of internal customs on the paper to communicate information beyond what the template allows for. Such examples are the recordings of more specific times in time slots (e.g. a performance in the ‘A’ afternoon row being specified as taking place between 3 to 5pm in Figure 1.1) and the circled red ‘L’s’ -- though we as outsiders can only speculate that this symbol means the only the lower level seats are available for this event, the lack of labelling indicates that those reading and writing in the ledger internally understand for sure what the symbol means. Not only are these customs clearly understood internally for those working at Carnegie, they are also resilient. This is evident from comparing ledgers at different points in time between 1955 to 2007; the circled red ‘L’s’ are present from the first volume in 1955 to the twenty-eighth volume in 1978 to the last volume in 2007 and so is the habit of writing more specific event times (see Figure 1.1-1.3). The fact that these internal practices passed through generations of employees to last more than 50 years show that recording and storing information on paper can set precedents that continue beyond the originator of those customs. The way that these customs must have been communicated between generations of employees to ensure consistency in how the ledger was kept is also another demonstration of the control and coordination in collecting and sorting information.

Figure 1.2

Figure 1.3

        Of particular interest is how susceptible to changes the ledger entries are; as previously noted, there are many time slots that have events that were crossed out and replaced with another event at least once, if not multiple times. However, it looks like the template purposefully leaves ample amounts of space. For the entries that required no modification, we can see that most events require no more than two lines of the ledger, yet all the time slot rows range from having three to seven lines (the amount of lines also changes depending on the time slot and whether it is a weekday or weekend). This could mean that the creator of the template anticipated that changes in scheduling would happen often and left enough space for edits to be made, also taking into account which days/times are most likely to undergo changes by giving those slots more lines. These ledgers also demonstrate a certain permanence associated with paper, as changes cannot be made without some visible form of evidence of the change (eraser marks, crossed out text, etc.).

The flexibility created for modifications in the ledgers is reminiscent of Linnaeus’s paper technologies discussed by Staffan Muller-Wille in “Natural history and information overload: The case of Linnaeus.”[2] Linnaeus initially had trouble inserting new information into bound manuscripts and eventually came up with systems that would allow him to amend his work as new information came to light as he knew that there was a significant possibility that the information that was recorded at the moment could change.[3] While Linnaeus used index cards and interleaved books, Carnegie Hall simply left more space for information to be crossed out and replaced. While Carnegie’s system seems less sophisticated, it does make more sense for it to be simpler because while there is an infinite amount of new information that Linnaeus could learn about a genus, Carnegie Hall’s schedule for a particular time slot at one of its venues is unlikely to change more than a few times given the book ledger’s purpose is to prevent scheduling mishaps. Therefore, predicting the amount of space that should be allotted for changes on a printed template is much easier for Carnegie compared to Linnaeus trying to foresee how much new information he would gather about a particular genus. A simpler system can also be attributed to differences between systems for an individual versus an organization. Linnaeus’s systems were for his own use, so he could add or modify information however he wanted as long as it made sense to him. Carnegie’s ledgers were for use across the entire organization, and thus needed to be readable to everyone and somewhat fixed.

The open-endedness of the booking ledgers and Linnaeus’s manuscripts run in contrast to most of the other artifacts. For example, the New York Philharmonic musician attendance book is very definitively static; there is little possibility that whether someone attended a rehearsal in the past will change in the future.[4] The same thinking applies to the meteorological journal of James Madison that was also shared: the weather of the past will not change.[5] This speaks to how different types of information must be recorded and stored differently despite the attendance book and meteorological journal also being paper artifacts. Most records collect present information that will always remain true in reference to the time of collection; however, the booking ledger collects information that pertains to the future and is therefore susceptible to change until the event date arrives and the event either happens or does not happen. Thus, when this kind of information is collected on paper, there must be flexibility created that accounts for the possibility of change.

The ledgers have already been digitized as they have been scanned and uploaded online. However, the information they contain is not readily accessible or searchable unless the searcher knows which exact volume and page to locate what they are looking for. Carnegie Hall has a performance history search that can be filtered by dates, performers, composers, works, and venues.[6] However, there are some discrepancies between what information is available for the search and the ledger itself. For instance, I did a search for the dates covered by the ledger in Figure 1.1. While the search turned up the performance in the Main Hall at 8:30pm for October 1, 1955 (Carmen Amaya), it only showed one performance in the Recital Hall for the same day despite four being scheduled on the ledger. In addition, since this tool is only for seeing the performance details, it does not include administrative information on the ledger such as the rate and close date. Other fields visible on the ledger but not captured in the performance search database include changes to the schedule, closing dates, and performance rates. Therefore, though this search provides a good starting point for events and events date, it does not include all the information we can see on the ledger.

To fully digitize this data and create a database similar to the performance history search, I would leverage the information available in this search. Ideally, through machine learning, it would be possible to identify different handwritings and extract the people involved in creating the ledger. This process would also transcribe the ledger text digitally. I would then cross-reference the digitized information from the machine learning technology with the performance history search database through the event/performer name and the event date. If the words from the event name in the search history and ledger contain matches and the dates match, then these are likely the same events. This would also give me exact event times as the search results have this information available, but the ledger entries do not always specify them within the parameters of morning, afternoon, and evening. From there, I would examine which events from the ledger have not matched an event from the performance search history and create new entries in the database with these events. The end result would be a database similar to the performance history search but with the administrative information from the ledgers incorporated.

Digitization would allow visualizations of information like changes in the historical performance rates at Carnegie Hall and how much ahead of time events are usually booked. Given the many internal customs that developed throughout decades of using these paper ledgers, a thorough digitization plan could also include interviewing someone who accessed the ledgers while they were in use to understand some of the unknowns on the ledgers (e.g. the meaning of the circled red ‘L’s’). Doing so would ensure that as many of these internal customs can be preserved during the digitization process, closing the gap between what can be conveyed on paper and what can be conveyed digitally.


[1] Carnegie Hall. (1955-2007). Carnegie Hall Booking Ledgers. Booking Ledgers Collection. Carnegie Hall Rose Archives. New York, NY, USA. https://collections.carnegiehall.org/Package/2RRM1TCSBJOS 

[2] Müller-Wille, Staffan, and Isabelle Charmantier. “Natural History and Information Overload: The Case of Linnaeus.” Studies in History and Philosophy of Science Part C: Studies in History and Philosophy of Biological and Biomedical Sciences 43, no. 1 (March 2012): 4–15. https://doi.org/10.1016/j.shpsc.2011.10.021.  

[3] Ibid.

[4] “New York Philharmonic Attendance Collection.” https://paperdata.benschmidt.org/exhibits/philharmonic 

[5] “James Madison's Meteorological Journal.” https://paperdata.benschmidt.org/exhibits/madison 

[6] Carnegie Hall. “Performance History Search.” https://www.carnegiehall.org/About/History/Performance-History-Search?q=&dex=prod_PHS