Coordinates of a Vector

What is a vector? While we often define a vector by a set of numbers like (x, y, z), this simple list of coordinates masks a much deeper and more powerful concept. A vector itself is a pure geometric object—an arrow in space—while its coordinates are merely a shadow it casts on a chosen set of reference axes. This distinction between an object and its description is one of the most fundamental ideas in mathematics and physics. Failing to grasp this difference can lead to confusion, while mastering it unlocks a unified perspective on the laws of nature.

This article navigates the concept of vector coordinates, starting from the foundational principles and building towards its most profound applications. In the "Principles and Mechanisms" chapter, we will dissect the relationship between a vector and its components, explore how to translate between different coordinate "languages," and introduce the crucial ideas of contravariant and covariant components essential for describing curved spaces. Following this, the "Applications and Interdisciplinary Connections" chapter will demonstrate how these transformation rules are not just mathematical formalities but a golden thread connecting fields as disparate as general relativity, quantum information theory, and chemistry. By journeying from simple Cartesian grids to the curved spacetime of Einstein's theories, you will learn to see beyond the numerical components and appreciate the invariant reality they represent.

Have you ever tried to give someone directions? You might say, "Go three blocks east and four blocks north." You've just given them a set of coordinates. But you could also have pointed and said, "Walk five blocks in that direction, straight towards the clock tower." The actual displacement, the arrow pointing from the start to the destination, is the same physical reality. Your descriptions, however, were completely different. This simple idea is the key to understanding vectors and their coordinates. A ​​vector​​ is a geometric object—an arrow with a magnitude and a direction—that exists independent of any description. Its ​​coordinates​​ are just a set of numbers we invent to label it, and these numbers depend entirely on the set of reference "rulers," or ​​basis vectors​​, we choose to measure it with.

The journey to understanding vectors is a journey in learning to distinguish the object from its shadow, the reality from its description. The true beauty of physics and mathematics is often found in discovering quantities that don't change when we change our point of view.

Most of the time, we live in the comfortable world of Cartesian coordinates. When we write a vector in $\mathbb{R}^3$ as $\mathbf{v} = \begin{pmatrix} 4 & -5 & 0 \end{pmatrix}^T$, we are implicitly saying it's "4 units along the x-axis, -5 units along the y-axis, and 0 units along the z-axis." These axes form our familiar, standard basis. But what if we want to change our language? What if, for a particular problem like creating a skewed perspective in a video game, a different set of basis vectors $\{\mathbf{b}_1, \mathbf{b}_2, \mathbf{b}_3\}$ is more natural?

The fundamental principle is that any vector $\mathbf{x}$ can be expressed as a unique combination of these new basis vectors:

The numbers $(c_1, c_2, c_3)$ are the coordinates of $\mathbf{x}$ in this new basis. How do we find them? We simply write out this vector equation in terms of the standard basis components and solve the resulting system of linear equations. It's a bit like being a cryptographer, translating a message from one code to another. For a given vector $\mathbf{x}$, finding its new coordinates $[ \mathbf{x} ]_{\mathcal{B}}$ is a standard exercise in this translation.

Conversely, if an ally tells you a vector's coordinates in their special basis, say $[ \mathbf{v} ]_{\mathcal{B}} = \begin{pmatrix} 2 & -1 & 3 \end{pmatrix}^T$, you can reconstruct the vector in the standard, common language by just calculating the linear combination $2\mathbf{b}_1 - \mathbf{b}_2 + 3\mathbf{b}_3$. The whole business is a two-way street.

And what about the simplest vector of all, the zero vector $\vec{0}$? You might notice that no matter how bizarre a basis you choose, its coordinates are always $(0, 0, \dots, 0)$. This isn't a coincidence. It's a direct consequence of the definition of a basis. The basis vectors must be ​​linearly independent​​, which means the only way the combination $c_1\mathbf{b}_1 + c_2\mathbf{b}_2 + \dots + c_n\mathbf{b}_n$ can equal the zero vector is if all the coefficients $c_i$ are zero. Any other possibility would mean one basis vector could be written in terms of the others, making it redundant—like having two different words for "east".

Solving systems of equations is work. Physicists, being cleverly lazy, are always on the lookout for a shortcut. Is there a "better" basis that makes life easier? Absolutely! Enter the ​​orthonormal basis​​. This is a set of basis vectors that are all mutually perpendicular (orthogonal) and have a length of one (normal).

When you use an orthonormal basis $\{\mathbf{u}_1, \mathbf{u}_2, \mathbf{u}_3\}$, finding the coordinates of a vector $\mathbf{v}$ becomes beautifully simple. No more solving systems of equations! The coordinate $c_i$ is just the dot product of the vector with the corresponding basis vector:

You can think of this as measuring the length of the shadow that $\mathbf{v}$ casts along the direction of $\mathbf{u}_i$. The dot product does this projection for you automatically. This trick is so powerful that it forms the foundation of many techniques in physics and engineering. In data science, for instance, Principal Component Analysis (PCA) is all about finding a special orthonormal basis that reveals the most significant directions of variation in a complex dataset, making the data much easier to understand. Choosing the right language doesn't just simplify the grammar; it can reveal the underlying story.

So far, our basis vectors have been fixed, unchanging arrows. But what happens if we move to a curved surface, like the Earth? The direction we call "east" is different in New York than it is in Tokyo. The basis vectors themselves change from point to point. This is the world of ​​general coordinate systems​​—like polar, cylindrical, or spherical coordinates—and the mathematics of curved spaces, known as differential geometry.

In this world, the simple change-of-basis matrix of linear algebra is no longer enough. The transformation of vector components from one coordinate system $(x^1, x^2)$ to another $(x'^1, x'^2)$ becomes dependent on where you are. The rule now involves a matrix of partial derivatives, the ​​Jacobian matrix​​:

This formula tells you how the numerical components $V^j$ of a vector field must change to precisely counteract the change in the local basis vectors, ensuring the vector itself remains the same geometric object. Whether you're transforming from Cartesian to polar coordinates for a physics problem or dealing with a more abstract non-linear coordinate change, this rule is the universal translator. The transformation factors are no longer constants; they are functions of position.

Here we arrive at a truly profound idea. In these general coordinate systems, there are two distinct, equally valid ways to describe a vector.

The components we've discussed so far, which transform using the Jacobian matrix as shown above, are called ​​contravariant components​​ (written with an upper index, $V^i$). They transform "contrary" to the basis vectors. Think of them as the familiar coefficients in a linear combination $\mathbf{V} = \sum_i V^i \mathbf{e}_i$.

But there's another way. We can define a set of ​​covariant components​​ (written with a lower index, $V_i$). These components describe how the vector interacts with the coordinate grid itself. Intuitively, you can think of them as measuring how many coordinate "level surfaces" the vector pierces. They transform using the inverse of the Jacobian matrix, "co-varying" with the basis vectors.

So, for a single vector $\mathbf{V}$, we have two different sets of numbers describing it! How are they related? The bridge between them is the most important object in geometry: the ​​metric tensor​​, $g_{ij}$. The metric tensor defines the very geometry of the space—it tells you how to calculate distances and angles. It acts as a dictionary to translate between the contravariant and covariant languages through a process called ​​raising and lowering indices​​:

For a diagonal metric, like in standard polar or spherical coordinates, this simplifies to multiplying each contravariant component by the corresponding diagonal entry of the metric tensor. This allows us to find the covariant components of a vector field if we know its contravariant ones, and vice-versa.

Why go through all this trouble of "upstairs" and "downstairs" indices? The payoff is immense. It allows us to express physical laws in a way that is completely independent of our chosen coordinate system.

Consider the dot product, or ​​scalar product​​, of two vectors, $\mathbf{P}$ and $\mathbf{Q}$. In simple Cartesian coordinates, it's just $P_x Q_x + P_y Q_y$. But how do we compute this in a wacky, curved coordinate system? The answer is breathtakingly elegant. The true, coordinate-independent scalar product is always found by "contracting" the contravariant components of one vector with the covariant components of the other:

This quantity, the sum $P^i Q_i$, is a ​​scalar invariant​​. No matter how you twist, stretch, or warp your coordinate system, its value remains the same. It represents a physical truth—like the projection of one vector onto another—that doesn't care about the language you use to describe it.

This exposes the deep importance of the formalism. What would happen if you were careless? What if you took the contravariant components of a vector $\mathbf{P}$ in one basis, and the covariant components of a vector $\mathbf{Q}$ in a different basis, and tried to multiply and add them? You would get a number, but this number would be pure gibberish. It would change whenever you changed your basis, and it would correspond to no physical reality. It would be a shadow of a shadow.

The distinction between covariant and contravariant components isn't just a notational game. It is the fundamental machinery that allows us to write down laws of nature—from electromagnetism to general relativity—that are universal, that are true for any observer in any reference frame. It's how we ensure we're talking about the vector, not just its fleeting shadow.

Now that we have grappled with the definition of a vector and its components, you might be tempted to ask, "So what?" Is this just a game for mathematicians, a scheme for re-labeling points and arrows? The answer, which is both startling and beautiful, is a resounding no. The rules for how a vector's components transform when we change our point of view are not merely a matter of bookkeeping. This transformation law is one of the most profound and unifying principles in all of science. It is the golden thread that ties together fluid dynamics, the curvature of spacetime, the bizarre world of quantum mechanics, and even the periodic table of chemical reactions. It is the physicist’s version of a universal grammar.

Let's begin our journey by looking at something familiar. Imagine a steady, uniform river flowing eastward. If we set up our coordinate system with the $x$-axis pointing east, the velocity vector of every water molecule is the same simple arrow, with components we might write as $(V_0, 0, 0)$. It's a beautifully simple description. But what if you are observing this river from a spinning merry-go-round in a park on the riverbank? To you, the water's motion seems incredibly complex. A drop of water that was heading straight now appears to be spiraling away. Expressed in your rotating, cylindrical coordinates, the velocity vector's components are no longer constant; they change depending on where the water is relative to you, involving sines and cosines of the angle. The physics hasn't changed—the river is still flowing steadily east—but our description has. The transformation rules allow us to translate between these two viewpoints and understand that they describe the same physical reality. The power of coordinates is not in finding the "right" one, but in understanding how to translate between all of them.

This idea is crucial whenever a problem has a natural symmetry. Describing the orbit of a planet in a rectangular grid is a nightmare of trigonometry. But in polar coordinates, centered on the sun, the description can become much simpler. If we have a vector described in a Cartesian grid, say with components $(V^x, V^y)$, the transformation laws give us a precise recipe to find its components $(V^r, V^\theta)$ in a polar grid. The vector—the physical quantity, like a force or a velocity—is the same. We have just projected its "shadow" onto a different set of axes.

Here we stumble upon a delightful surprise. What, you might ask, are the components of the position vector itself—the arrow pointing from the origin to a point in space? In Cartesian coordinates $(x, y, z)$, the answer seems obvious: the components are the coordinates, $(x, y, z)$. But don't be fooled! This is a special property of this one particular coordinate system. If we ask the same question in spherical coordinates $(r, \theta, \phi)$, the answer is not $(r, \theta, \phi)$. After applying the rules of transformation, we find the components are simply $(r, 0, 0)$. At first, this seems absurd. But think about it. The position vector points from the origin outwards. That is exactly the direction of the radial basis vector. So, it has a component of size $r$ in the radial direction, and zero component in the angular directions. This little paradox forces us to appreciate the true meaning of components: they are projections onto a local set of basis vectors, which themselves can point in different directions at different locations.

The real purpose of wrestling with these transformations is to find what doesn't change. We call these things "invariants." The most fundamental invariant of a vector is its length. The components can twist and turn, get larger or smaller, but they must conspire in such a way that the vector's length remains the same. Consider a simple reflection across the $xy$-plane. A vector's $z$-component flips its sign, but its length-squared, $v_x^2 + v_y^2 + v_z^2$, is unchanged because $(-v_z)^2 = v_z^2$. This principle is the heart of physics. Physical laws must be invariant statements, true in any valid coordinate system.

As we venture into more general "curvilinear" coordinates, where the axes can be curved and not perpendicular, a new subtlety appears. We discover that there are two natural ways to define components. We can have "contravariant" components, which are found by projecting the vector parallel to the coordinate axes onto the basis vectors, and "covariant" components, which are related to projections perpendicular to coordinate surfaces. For simple Cartesian systems, they are identical. But for polar or spherical coordinates, they are different beasts. To get from one to the other, or to calculate a simple invariant like the vector's magnitude, we need a special dictionary. This dictionary is an object called the "metric tensor," which encodes all the geometric information of our coordinate system—the local lengths and angles. It is what allows us to compute the true length of a vector from its components, no matter how strange our coordinates are.

With this powerful toolkit, we can leave the comfort of flat paper and venture onto curved surfaces. Imagine you are an ant living on a giant, bumpy potato. Your world is two-dimensional and curved. Your velocity vector as you crawl along the surface must always lie tangent to the surface. It doesn't point out into the third dimension, which you don't even know exists. The components of your velocity are not with respect to some external $(x, y, z)$ axes, but with respect to the local coordinate grid you've drawn on your potato world. This is the essence of differential geometry, the language Einstein used to describe a universe where spacetime itself is a curved surface.

And this brings us to relativity. One of the central ideas of Einstein's theory is that the laws of physics should look the same for all observers. Let's see what this means for our vectors. In the flat spacetime of special relativity, a stationary observer sees time flowing forwards. This physical concept, "time translation," can be represented by a simple Killing vector with components $(1, 0, 0, 0)$. Now, consider an observer in a rocket, accelerating relentlessly through space. Their spacetime is described by so-called Rindler coordinates. When we translate the simple time-translation vector into the rocket's frame, it becomes a complicated, position-dependent vector field. This isn't just a mathematical trick. It has profound physical consequences. This transformation is at the root of the Unruh effect, a startling prediction that the accelerating observer will feel heat and see particles in what the stationary observer calls a perfect, cold vacuum. The physics is contained entirely in the transformation law of the vector components.

The power of this abstraction—an object defined by how its components transform—is so immense that it has been co-opted by fields far from geometry. In quantum information theory, the state of a single qubit can be visualized as a vector, the "Bloch vector," pointing somewhere inside a sphere. More generally, any process or measurement can be represented by an operator, and this operator can be decomposed into components using the Pauli matrices as a basis, forming a "Pauli vector". When a qubit interacts with its environment and loses information—a process called decoherence—the evolution is perfectly described by how this abstract vector shrinks and rotates. The physics of quantum noise becomes the geometry of a vector transformation.

This way of thinking, through symmetry and transformations, is also the language of modern chemistry. A molecule like ammonia has a certain geometric symmetry (a three-fold rotation and three reflection planes, a group called $C_{3v}$). Whether the molecule can absorb light or participate in a certain type of spectroscopic process, like Raman scattering, depends on how quantities like the electric dipole or the polarizability tensor transform under these symmetry operations. By analyzing the transformation rules for the components of the polarizability tensor, chemists can derive "selection rules" that predict which molecular vibrations will be visible in their experiments and with what light polarization. The symmetry of the component transformations dictates what is allowed and what is forbidden by nature.

Finally, at the very deepest levels of theoretical physics, this idea of a vector and its components reveals itself as part of a yet grander, more mysterious structure. In the rarefied world of string theory, physicists study rotations in eight dimensions. Here, the group of rotations, $SO(8)$, exhibits an exceptional property called "triality." It turns out that in eight dimensions, there are not one, but three fundamentally different types of 8-component objects. There is the familiar vector, and two other entities known as spinors. Triality is a profound symmetry that can mix and exchange these three objects—the vector can be turned into a spinor, and a spinor into a vector. The humble vector we started with is revealed to be just one member of a holy trinity, hinting at a deep, hidden coherence in the mathematical fabric of reality.

And so, we have come full circle. We began with the simple, almost mundane question of how to write down the components of an arrow. By following this question relentlessly, we discovered a universal principle that governs the description of physical laws. The transformation of coordinates is the key that unlocks the geometry of curved spacetime, the dynamics of quantum states, and the fundamental symmetries of nature. It teaches us to distinguish the arbitrary description from the invariant reality. In this simple idea, we find one of the most elegant and powerful manifestations of the unity of science.