In Article I (Re: The Library), I laid the groundwork for something most of us already know instinctively: The library isn’t an add-on to orchestral life. It’s part of the artistry and the infrastructure that quietly holds everything together. That infrastructure—the preparation, decisions, and labor that exist long before anything happens onstage—is easy to overlook precisely because it usually works. When the infrastructure is healthy, it’s invisible. You don’t think about it because you don’t have to; you simply trust it to do what it was built to do and move through your day assuming it’s quietly working in the background.

What no one sees (unless they’re behind the scenes or part of the machinery making it happen day to day) is how that same infrastructure is holding up under the weight and pressure it carries every single day. The tiny frictions. The wear and tear. The compounding strain of one more request, one more late change, one more unexpected need—all small pressures that quietly add weight to a system already operating at capacity. Because most people never feel that growing strain directly, they don’t recognize where it’s building, but the few of us who are inside that infrastructure and part of its structure can see exactly where it’s pressing the hardest and where it’s starting to buckle. 

For an even clearer picture, think about a levee system during a storm. When a storm hits, water levels rise and pressure builds steadily against the barriers that comprise them. Nothing looks wrong. The levees are doing what they are supposed to do: holding. On the surface, everything still appears stable, even as the force against them increases. What you don’t see, though, is that the wall isn’t equally strong everywhere. Over time, from previous storms, some sections have taken more hits than others. Some have been patched instead of reinforced. Some have been rebuilt. Some have been eroded completely and never repaired, carrying more strain than anyone realized. So as the pressure keeps increasing, it isn’t pressing against a perfectly even structure. Instead, it’s pressing hardest on the spots that have already been worn thin. Those weaker areas give way first, and eventually a breach follows.

When that breach happens, it often gets talked about like bad luck. A rough patch. A one-off emergency that just needs a quick fix. But, breaches aren’t random, and they aren’t about luck. They’re the visible result of strain that’s been accumulating for a long time due to pressure rising faster than maintenance and reinforcement. Once the load exceeds what the wall was built—and updated—to handle, one crack becomes many, exposing a system that’s been quietly carrying more than it was ever designed, staffed, or resourced to sustain. That’s how pressure reveals where a system has been stretched past its limits—and that same pattern shows up in orchestras.

In an orchestra, the collective bargaining agreement functions much like that levee wall. It’s the structure designed to absorb pressure, distribute strain, and keep any one part of the system from taking more than it can sustain. When the language is clear and actually upheld, it holds because the load is fairly balanced across the whole unit. However, the moment the language leaves work unnamed and open to interpretation, or the protections in it stop being enforced consistently, the balance shifts and the pressure begins to concentrate on the parts of the system with the least protection.

Since this series focuses on the library, it’s worth looking at how the above dynamics show up there.

When bargained language exists but stops being honored consistently and starts getting treated like it has room to bend, a “just this once” makes the week work, or that tour work, or that really special program with that amazing artist work. Then comes another. And another. Bit by bit, those exceptions become the norm and begin reshaping the agreement into something that was never actually bargained at the table. And eventually, the cost of that reshaping falls to the people who rarely benefit from the thing those “just this once” exceptions were traded for to pay, widening the gap between who is protected and who is expected to absorb the strain.

Conversely, when language is unclear or missing altogether—as is often the case with the library in CBAs—the work still has to happen. The orchestra still needs the same preparation, decisions, and labor to make the week work, so the gap between what’s required and what’s defined gets filled however the moment demands. Pages are bowed, parts are corrected, materials are assembled, questions get answered. But, because the contract doesn’t clearly define and account for that labor, the library’s work is easy to overlook. Once that essential labor is easily overlooked, it becomes much easier for it to be pushed around, handed off, or expected without the same clarity or safeguards other work inside the bargaining unit receives.

To be fair, playing musicians don’t have the day-to-day details of their jobs written into the contract either. However, that specific labor is still structurally anticipated and protected through service definitions, limits, and other carefully negotiated safeguards. Even so, playing musicians are no strangers to having their work misunderstood due to much of it happening out of sight. A good portion of the week-in, week-out labor is invisible: individual practice, score and part study, physical maintenance, instrument care—all of that happening long before everyone comes together to rehearse and perform. That invisibility creates space for a different narrative to be introduced: “Musicians only work 20 hours a week.” 

This phrase, usually heard around the negotiation process, is used to collapse invisible labor—practice, preparation, recovery, and cumulative physical strain—into something that looks part-time on paper. That framing is dangerous because it makes it easier to undervalue the work, justify paying less, or argue against paying more. Contracts exist to push back against the practice of undermining musician labor. In theory, they are meant to account for the reality of the work and counter the rhetoric used to shrink it. In practice, though, they don’t always do that evenly across the bargaining unit.

That’s exactly what’s happening when it comes to librarians and the labor they perform. They remain clearly misunderstood within the very system their work makes possible, which makes it easier for both the work and the people doing it to be undervalued, underprotected, improperly classified, or quietly left out of the standard language altogether. 

Library work is well defined. It’s just rarely described accurately by everyone (except the librarians). Everyone knows the library is critical, but far fewer people can clearly articulate its role because they only encounter the work through a vague sense that the music “gets ready.” Library work is precise, technical, and constant, yet it’s often talked about as if it’s a single, simple step in the process. When work gets reduced like that, people start making assumptions about it—how long it takes, what it requires, how many people are needed to do it, and who is responsible for it. Those assumptions shape expectations, timelines, and demands placed on the library before anyone checks whether they’re realistic (except the librarians—they’ve been saying it’s unrealistic from the beginning, so it isn’t exactly news to them). This is one of the first signs of stress along the wall.

Another stress point shows up in information: in who people think has it, who is withholding it, and what happens when others act on what they assume they know. Inside an orchestra, information behaves like currency—traded casually, hoarded strategically, and spent freely, while the work of verifying it sits elsewhere. Because the library is seen as the place that “just knows,” it becomes an expectation that every piece of information gets funneled there to be confirmed. Someone hears something from another ensemble, remembers something vaguely from years ago, or simply believes a thing must exist because it should and suddenly that belief arrives at the library’s door as a task, a directive, or a deadline.

When the library doesn’t already have the answer, or when the answer requires time, research, or confirmation, the reaction often isn’t curiosity or partnership. It’s frustration. The library starts to be seen as “difficult” or “slow,” not because the request itself has been examined for unreasonableness, but because the assumption behind it was never examined at all. At that point, the labor shifts and instead of preparing music, librarians are now having to repair the request itself by clarifying what is true, what is false, what needs research, and what never existed in the first place. Additionally, because orchestras talk to one another, misinformation travels quickly. A mistaken belief in one hall becomes an urgent demand in another long before anyone verifies it. Librarians end up searching for materials that never existed, apologizing for things that were never real, and untangling certainty that turns out to be wrong.

These are just some of the ways pressure accumulates along those stress points long before anyone names it. Hours that should be spent preparing music disappear into explaining our jobs, arguing for structural protections, performing labor that was never formally accounted for, and validating assumptions that evaporate under scrutiny. Work lands on the library based on partial information or unexamined belief, and the time, labor, and emotional fallout of sorting it out settles there as well. When expectations arrive assumed rather than verified, the pressure to clarify what is actually possible pushes straight into the library, compounding already tight timelines, limited staffing, and the thin margins under which the work is getting done.

This is where accumulated pressure turns into risk, which is the stage just before visible failures begin. On top of resolving assumptions, the library is simultaneously forced to manage the risk created when attention drops away from critical work—all in real time. As mentioned earlier, the library operates under hard, immovable deadlines in service of both the organization onstage and the work happening behind the scenes. Under duress, the solutions tend to look the same: more bodies, found quickly, pulled from the closest available pool, and justified by urgency. The goal in that moment isn’t elegance, fairness, longevity, or parity. It’s a single idée fixe: avoiding failure so the show can go on.

This is the environment in which library substitutes, interns, and additional staff positions are brought in (with varying compensation levels, if they are compensated at all) and end up carrying real responsibility. It’s how people with experience, or people who need more of it, say yes to work that looks like opportunity even when it’s really coverage. (There is no degree for performance librarianship. The only way to gain experience is to be in a library, learning the job by doing it.) Over time, unique, unpaid, or underpaid help gets framed as temporary, educational, or exceptional—even as it becomes routine and, eventually, a regular fixture in the library. I’ve seen this pattern up close, from more than one vantage point. Early-career library labor is often treated as a proving ground, where real responsibility is assigned without corresponding protection or pay, under the assumption that experience itself is compensation—it isn’t. And the thing is, that logic doesn’t magically disappear once someone has more experience or authority. It just gets normalized.

What matters here isn’t intent, but repetition. When workarounds like these repeat, they stop being exceptions and start setting expectations—not because anyone formally decided to change the rules, but because the system quietly adjusted to what it was already relying on. Short-term fixes made under pressure become standing arrangements, and once that happens, the effects start shaping the structure itself. The workaround becomes the rule, showing up both in the CBA and in how its language is interpreted and applied. Compensation is where the “more bodies for less money” becomes impossible to pretend is harmless.

As everyone is aware, there’s a difference between discretionary pay—overscale, premiums, negotiated extras—and base pay. Base pay is the floor. It’s what the agreement guarantees to everyone it covers. Or does it? A quick look at library language in many CBAs, or at ICSOM wage charts, shows that this base is not, in fact, guaranteed to all librarians. When musicians in the same bargaining unit (and more specifically, in the same section) are paid less than the guaranteed base pay, that isn’t an individual agreement issue. That’s a structural issue, and it’s where the library’s treatment inside the bargaining unit becomes impossible to ignore.

In many CBAs, librarians are technically included in the unit while being paid as a percentage of base scale, depending on the number of people in the section, rather than receiving the guaranteed base itself. This arrangement is often framed as a compromise, a way of “bringing the library in,” but functionally, it creates a tiered system inside the unit: different base pay for the same work, limited or nonexistent access to overscale in some cases despite holding titles or endowed chairs, and a compensation ceiling that cannot be crossed no matter how much responsibility the job actually carries. The result isn’t just lower pay. It’s a structural divide. Librarians working side by side—performing the same core functions and carrying the same week-to-week responsibility—are compensated differently by design. Over time, that design difference morphs into a tiered system within the library itself—an internal hierarchy where people doing the same work are assigned different standing and base pay by structure.

Once that logic is accepted inside the bargaining unit at the library level, what started as a stress fracture in one section becomes a crack in the CBA itself, affecting not only the library but the entire bargaining unit. It establishes a precedent—in both function and interpretation—that base pay is negotiable by classification, that full participation in the unit can come with carve-outs, and that essential work can be discounted without technically violating the agreement, because the agreement already made room for it. At that point, the compensation floor isn’t holding, and once that floor erodes, other protections tied to the CBA—including time off, overtime, relief, and sustainable scheduling—begin to erode alongside it.

This is how that erosion shows up in practice:

Substitutes and extras are paid below base scale—or brought in for a full week of work but paid only per service or per task, rather than being guaranteed a full week’s base pay. Early career musicians—particularly at the conservatory level and beyond—are asked to perform covered work without pay, or for so-called “educational rates.” “Pipeline” or “emerging” labels get attached to professional responsibility, offering a rationale for paying less for the same work.

Erosion doesn’t stop at pay. In wind sections, weeks where a player is “not needed due to instrumentation” get counted as relief weeks (even though relief is clearly defined in the CBA), effectively conflating relief with rotation and denying players the contractual rest those weeks are meant to provide. Doubling clauses tell a similar story: Bargained protections get bypassed through individual agreements, where players  are paid a small amount of overscale to cover doubling that the CBA explicitly defines and compensates at a higher rate. On paper, it looks like flexibility. In practice, it undercuts mandatory language and shifts cost and fatigue back onto the player.

There’s a belief that workarounds are temporary. That’s a fallacy. Discounted labor, blurred definitions, and one-off “solutions” get repeated, relied on, and quietly absorbed into the system. Over time, reduced pay and reduced protections stop looking like exceptions and start functioning like categories. That’s how precedent gets established. Not all at once, but through repetition that slowly reshapes what the agreement is understood to allow.

What becomes clear from these patterns is that what starts as bargained “solutions” for the library—more bodies doing the same work for less money, members of the same section not making guaranteed base pay, amended or reduced time off—doesn’t stay in the library. It moves through the entire infrastructure of the bargaining agreement, pressing against other weak points in the contract. As pressure builds unevenly across the agreement, breaches begin to appear. Slowly at first, and often in places that are easy to miss. Systems rarely fail all at once. They fail the way levee systems do.

I recognize this is a lot to sit with. Systems are often invisible to the people least burdened by them, so it can feel uncomfortable when someone reveals how they’re actually working (or in this case, not working). So what does repair actually require? Structural change. Precedent has to be interrupted where it first formed—and where it’s still being reinforced. This means repairing the contract at the library level. It requires real language that reflects the work as it actually exists, names the labor the system already relies on quietly, creates protections around that labor, and makes them enforceable instead of optional.

Next up: Finishing the series—Re: Design