The Standard Dot Product On R^n

The Standard Dot Product

Vectors in \(\mathbb{R}^n\) have coordinates, and the standard dot product combines those coordinates into one real number. This number is useful both for calculating lengths and for comparing how two vectors point in relation to one another. Its definition is short, but its properties matter: the standard dot product is the basic example of a real inner product.

Write vectors as columns, so \(\mathbf{u}=(u_1,\dots,u_n)^{\top}\) and \(\mathbf{v}=(v_1,\dots,v_n)^{\top}\), where each coordinate is real. The dot product multiplies corresponding coordinates and adds the results. In matrix notation, it can also be calculated by multiplying the row \(\mathbf{u}^{\top}\) by the column \(\mathbf{v}\).

The dot product is a real number, not a vector. The length is also a real number and is always nonnegative, because it is the square root of a sum of squares. For example, in \(\mathbb{R}^3\), if \(\mathbf{u}=(u_1,u_2,u_3)^{\top}\) and \(\mathbf{v}=(v_1,v_2,v_3)^{\top}\), then \(\mathbf{u}\cdot\mathbf{v}=u_1v_1+u_2v_2+u_3v_3\), while \(\|\mathbf{u}\|=\sqrt{u_1^2+u_2^2+u_3^2}\).

Computing Dot Products and Lengths

The same calculation applies in any \(\mathbb{R}^n\): the number of coordinates determines how many products are added, but the rule does not change. It can be helpful to write out every coordinate product before combining them, especially when vectors contain negative entries.

Why the Standard Dot Product Is an Inner Product

A real inner product on a real vector space is a rule that assigns a real number to each pair of vectors and satisfies symmetry, linearity in each argument, and positive definiteness. Here, positive definiteness means that a vector’s inner product with itself is nonnegative, and equals zero exactly for the zero vector. The next theorem verifies these properties directly for the standard dot product on \(\mathbb{R}^n\).

Proof. Write \(\mathbf{u}=(u_1,\dots,u_n)^{\top}\), \(\mathbf{v}=(v_1,\dots,v_n)^{\top}\), and \(\mathbf{w}=(w_1,\dots,w_n)^{\top}\). For symmetry, real-number multiplication is commutative, so \(\mathbf{u}\cdot\mathbf{v}=\sum_{i=1}^{n}u_iv_i=\sum_{i=1}^{n}v_iu_i=\mathbf{v}\cdot\mathbf{u}\). The \(i\)th coordinate of \(a\mathbf{u}+b\mathbf{v}\) is \(au_i+bv_i\). Therefore \((a\mathbf{u}+b\mathbf{v})\cdot\mathbf{w}=\sum_{i=1}^{n}(au_i+bv_i)w_i =a\sum_{i=1}^{n}u_iw_i+b\sum_{i=1}^{n}v_iw_i =a(\mathbf{u}\cdot\mathbf{w})+b(\mathbf{v}\cdot\mathbf{w})\). For linearity in the second argument, the \(i\)th coordinate of \(a\mathbf{v}+b\mathbf{w}\) is \(av_i+bw_i\), and hence \(\mathbf{u}\cdot(a\mathbf{v}+b\mathbf{w})=\sum_{i=1}^{n}u_i(av_i+bw_i) =a\sum_{i=1}^{n}u_iv_i+b\sum_{i=1}^{n}u_iw_i =a(\mathbf{u}\cdot\mathbf{v})+b(\mathbf{u}\cdot\mathbf{w})\). Finally, \(\mathbf{u}\cdot\mathbf{u}=\sum_{i=1}^{n}u_i^2\), a sum of nonnegative real numbers. It equals zero if \(\mathbf{u}=\mathbf{0}\). Conversely, if the sum is zero, no \(u_i^2\) can be positive, since then the whole sum would be positive. Thus every \(u_i^2=0\), so every \(u_i=0\), and \(\mathbf{u}=\mathbf{0}\). This proves all the stated properties. \(\square\)

Symmetry says the order of the two vectors does not affect the answer. Linearity says that combining vectors in either argument distributes through the dot product and allows scalars to be factored out. These properties are not extra computational rules invented for convenience: they follow from the coordinate formula and the ordinary rules of real arithmetic.

The Dot Product as Signed Coordinate Alignment

Each product \(u_iv_i\) compares the two coordinates in position \(i\). If the coordinates have the same sign, their product is positive; if they have opposite signs, it is negative; and if either is zero, the product is zero. The full dot product adds these signed contributions. A positive result means positive contributions outweigh negative ones, while a negative result means negative contributions outweigh positive ones. A zero result can occur through cancellation, as in the preceding example.

The numerical value also depends on the sizes of the vectors. For example, multiplying one vector by a positive scalar multiplies the dot product by that scalar, so the value is not a scale-free comparison of direction by itself. The scaling rule follows directly from linearity.

Proof. The \(i\)th coordinates of \(a\mathbf{u}\) and \(b\mathbf{v}\) are \(au_i\) and \(bv_i\). By the definition of the standard dot product, \((a\mathbf{u})\cdot(b\mathbf{v})=\sum_{i=1}^{n}(au_i)(bv_i) =ab\sum_{i=1}^{n}u_iv_i=ab(\mathbf{u}\cdot\mathbf{v})\). This computation also covers \(a=0\) or \(b=0\), since both sides are then zero. \(\square\)

Consequently, changing the sign of exactly one vector reverses the sign of the dot product, while multiplying both vectors by negative scalars preserves its sign. Dot products are particularly useful when the vector lengths are fixed or when the sign itself is the question. To interpret the value as an angle or to compare it independently of length requires additional results, which are not needed for coordinate computations.

Common Mistakes and a Coordinate Check

A frequent error is to lose a negative sign when multiplying corresponding coordinates. For the vectors in the first example, a mistaken calculation might replace \((-1)(4)\) by \(+4\), giving \(2+4-6=0\). That answer is incorrect: the second coordinate product is \(-4\), so the correct calculation is \(2-4-6=-8\). A reliable check is to write each product with parentheses before adding.

Another mistake is to confuse the dot product with the length. The dot product \(\mathbf{u}\cdot\mathbf{v}\) uses two vectors and may be negative; the length \(\|\mathbf{u}\|\) uses one vector twice and cannot be negative. For any \(\mathbf{u}\), the formulas make the distinction explicit:

The standard basis offers a compact check of the coordinate rule. The vector \(\mathbf{e}_i\) has a \(1\) in coordinate \(i\) and zeros elsewhere. Thus \(\mathbf{e}_i\cdot\mathbf{e}_j=1\) when \(i=j\), and \(0\) when \(i\ne j\): when the positions differ, every coordinate product is zero. More generally, writing \(\mathbf{u}=\sum_{i=1}^{n}u_i\mathbf{e}_i\) and \(\mathbf{v}=\sum_{i=1}^{n}v_i\mathbf{e}_i\) gives \(\mathbf{u}\cdot\mathbf{v}=\sum_{i=1}^{n}u_iv_i\), since only matching basis positions contribute.