The Geometry and Analysis of Inverse Trigonometric Functions

When working with inverse trigonometric functions in symbolic mathematics, one quickly encounters puzzling inconsistencies. Why does Mathematica simplify sin(arccos(x)) to √(1-x²) while leaving sinh(arccosh(x)) as √(x²-1)? This apparent asymmetry between trigonometric and hyperbolic functions reveals deep connections between geometry, complex analysis, and the conventions we use to define these functions.

The Geometric Foundation

Let's start with the simplest case. For real numbers x between 0 and 1, we can construct a right triangle with hypotenuse length 1 and one leg of length x. The angle θ between these two sides satisfies cos(θ) = x, meaning arccos(x) = θ. By the Pythagorean theorem, the remaining leg has length √(1-x²), which is exactly sin(θ). Therefore, sin(arccos(x)) = √(1-x²) for 0 < x < 1.

This geometric argument is elegant and intuitive, but it only covers a limited domain. To extend these identities to all real numbers and eventually to complex numbers, we need more sophisticated tools.

Analytic Continuation and Branch Cuts

The key to extending inverse trigonometric functions is analytic continuation. We can define arccos(z) for any complex number z by making branch cuts along the real intervals (-∞, -1] and [1, ∞). These branch cuts are necessary because cosine is not one-to-one on the entire complex plane.

When extending a function across a branch cut, we must choose a convention. The standard choice is to make arccos(z) continuous when approaching the branch cut from above (from the second quadrant). This means that as we approach a point on the branch cut from slightly above the real axis, the function's value approaches the same limit as if we were approaching from the second quadrant.

The Square Root Function's Role

Here's where things get interesting. The square root function also requires a branch cut, typically chosen along (-∞, 0]. When we write √(1-z²), we're using the principal branch of the square root function, which is continuous when approaching the branch cut from above.

This shared convention between arccos and square root is crucial. Since sin(arccos(x)) = √(1-x²) for real numbers between -1 and 1, and both sides are analytic functions that agree on a set with a limit point, they must be equal everywhere in their common domain by the identity theorem from complex analysis.

The Critical Test: Branch Cut Consistency

The final verification is whether the two expressions agree along the branch cuts themselves. Since both arccos and the square root function are defined by continuous extension from the second quadrant, their composition should maintain this continuity. This consistency check confirms that sin(arccos(z)) = √(1-z²) holds for all complex z, not just real numbers.

Why Hyperbolic Functions Behave Differently

This brings us back to our original puzzle. The hyperbolic functions sinh and cosh don't share the same geometric intuition as their circular counterparts. While we can visualize sin and cos using right triangles, there's no analogous geometric construction for sinh and cosh that naturally suggests the same simplification patterns.

Moreover, the branch cut structure for inverse hyperbolic functions differs from inverse trigonometric functions, leading to different conventions and behaviors in symbolic computation systems.

Implications for Symbolic Computation

The behavior of these functions in computer algebra systems reflects deeper mathematical principles. When a system like Mathematica or SymPy simplifies an expression, it's applying a combination of:

• Known identities and transformations
• Branch cut conventions
• Domain assumptions
• Heuristic patterns

Understanding these underlying principles helps us interpret the results and sometimes guide the system toward the desired simplification.

Beyond the Basics

This exploration reveals a broader lesson about mathematical functions: their behavior depends not just on their definition but on how we choose to extend them beyond their natural domain. The same function can have different representations depending on whether we're working with real numbers, complex numbers, or specific branches of multi-valued functions.

For instance, the identity sin(arccos(z)) = √(1-z²) holds for all complex z when using the principal branches, but would fail if we chose different branch conventions. This sensitivity to conventions is a recurring theme in complex analysis and has practical implications for numerical computation and symbolic manipulation.

Conclusion

The seemingly simple question of why sin(arccos(x)) simplifies while sinh(arccosh(x)) doesn't opens a window into the rich interplay between geometry, analysis, and computation. It reminds us that mathematical expressions carry implicit assumptions about domains, branches, and conventions that can dramatically affect their behavior.

For practitioners working with symbolic mathematics, this understanding provides both practical guidance and philosophical insight. We learn to appreciate why certain simplifications work and others don't, and we gain a deeper appreciation for the careful choices that underlie even the most basic mathematical operations.

The next time you encounter an unexpected result from a computer algebra system, remember that it might be revealing not a bug, but a subtle feature of the mathematical structures we're working with—structures that connect ancient geometric intuition to modern complex analysis in ways that continue to surprise and enlighten us.