What Scalar Multiplication Does
A matrix records several numbers in a fixed rectangular arrangement. Multiplying a matrix by a scalar changes the numbers in that arrangement: the same scalar is applied to every entry. The size of the matrix does not change. This operation is called scalar multiplication, and its entry-by-entry definition makes it possible to compute positive, negative, fractional, and zero scalar multiples without ambiguity.
In this tutorial, scalars are real numbers. Suppose \(\mathbf{A}\) has \(m\) rows and \(n\) columns, with entry \(a_{ij}\) in row \(i\), column \(j\). The scalar \(c\) is one real number, used for every entry of \(\mathbf{A}\). It is not a matrix product: no rows are paired with columns, and there is no dimension-matching condition beyond knowing the size of \(\mathbf{A}\).
The notation \(c\mathbf{A}\) puts the scalar before the matrix. For example, if \(c=4\), every entry is multiplied by \(4\); if \(c=-1\), every entry changes sign; and if \(c=0\), every entry becomes zero. These are consequences of using the same real-number multiplication on each entry. The matrix retains its original row and column counts in all three cases.
To calculate a scalar multiple reliably, identify the scalar first, then multiply each entry separately. Keep the locations fixed: the entry in row \(i\), column \(j\) of the result comes from the entry in row \(i\), column \(j\) of the original matrix. A useful check is to compare the dimensions and then revisit every entry, especially when the scalar is negative.
Computing Positive and Negative Scalar Multiples
Worked Example: A Positive Scalar and a Rectangular Matrix
Find \(3\mathbf{A}\) for \(\mathbf{A}=\begin{bmatrix}2&-1&0\\4&3&-2\end{bmatrix}\).
The matrix has \(2\) rows and \(3\) columns. Multiply each of its six entries by \(3\), keeping each result in its original position:
Check: The result still has \(2\) rows and \(3\) columns. Its entries, in order by row, are \(6,-3,0,12,9,-6\), which are respectively \(3\cdot2\), \(3\cdot(-1)\), \(3\cdot0\), \(3\cdot4\), \(3\cdot3\), and \(3\cdot(-2)\). Every entry satisfies the definition.
A negative scalar multiplies every entry, not only entries that were negative. The sign of each product follows ordinary real-number multiplication: a negative scalar times a positive entry gives a negative result, a negative scalar times a negative entry gives a positive result, and a zero entry remains zero. It is worth writing the products before simplifying when several signs appear in the same matrix.
Worked Example: Multiplying by a Negative Scalar
Compute \(-2\mathbf{B}\) for \(\mathbf{B}=\begin{bmatrix}-3&2\\0&-1\\4&-5\end{bmatrix}\).
Apply \(-2\) to all six entries:
Check: The original matrix and the result both have \(3\) rows and \(2\) columns. Entry by entry, \(6=(-2)(-3)\), \(-4=(-2)(2)\), \(0=(-2)(0)\), \(2=(-2)(-1)\), \(-8=(-2)(4)\), and \(10=(-2)(-5)\). In particular, both negative entries of \(\mathbf{B}\) become positive, while its positive entries become negative.
Fractions are handled in exactly the same way. Each entry is multiplied by the fraction, and the product can then be simplified. There is no requirement that the resulting entries be integers. The calculation is still entrywise, and the dimensions are unchanged.
Worked Example: A Fractional Scalar
Find \(-\tfrac12\mathbf{C}\) for \(\mathbf{C}=\begin{bmatrix}4&-6&2\\-8&0&10\end{bmatrix}\).
Multiply every entry by \(-\tfrac12\) and simplify each product:
Check: Both matrices have \(2\) rows and \(3\) columns. The six entries check directly: \(-2=(-\tfrac12)(4)\), \(3=(-\tfrac12)(-6)\), \(-1=(-\tfrac12)(2)\), \(4=(-\tfrac12)(-8)\), \(0=(-\tfrac12)(0)\), and \(-5=(-\tfrac12)(10)\).
Zero Scalars and Zero Matrices
The zero scalar needs no special computational method: it is multiplied by each entry. The result is a matrix all of whose entries are zero. For a particular size, this is the zero matrix of that size. The dimensions still matter; for example, a \(2\times3\) zero matrix and a \(3\times2\) zero matrix do not have the same size.
Proof. Write \(\mathbf{A}=[a_{ij}]\). By definition, the entry in position \((i,j)\) of \(0\mathbf{A}\) is \(0a_{ij}=0\). This holds for every \(i\) and \(j\), so \(0\mathbf{A}\) has the same dimensions as \(\mathbf{A}\) and every entry is zero. Now suppose \(c\ne0\). If \(\mathbf{A}\) is the zero matrix, then each \(a_{ij}=0\), and each entry of \(c\mathbf{A}\) is \(ca_{ij}=c(0)=0\). Conversely, if \(c\mathbf{A}\) is the zero matrix, then \(ca_{ij}=0\) for every \(i,j\). Since \(c\ne0\), division by \(c\) gives \(a_{ij}=0\) for every \(i,j\), so \(\mathbf{A}\) is the zero matrix. \(\square\)
Worked Example: The Zero Scalar
Compute \(0\mathbf{D}\) for \(\mathbf{D}=\begin{bmatrix}5&-2\\1&0\\-3&4\end{bmatrix}\).
Multiply each entry by zero:
Check: The result is a \(3\times2\) matrix, as required. Each position contains zero: \(0(5)=0\), \(0(-2)=0\), \(0(1)=0\), \(0(0)=0\), \(0(-3)=0\), and \(0(4)=0\).
What a Nonzero Scalar Changes
When the scalar is nonzero, it cannot turn a nonzero entry into zero: if \(ca_{ij}=0\) and \(c\ne0\), then \(a_{ij}=0\). Thus the positions of the zero entries stay fixed under multiplication by a nonzero scalar. The signs can be described precisely as well. A positive scalar preserves the sign of each nonzero entry, whereas a negative scalar reverses it. This statement concerns individual entries; it does not say that a matrix has a single overall sign.
Proof. By definition, \((c\mathbf{A})_{ij}=ca_{ij}\). If \(a_{ij}=0\), then \(ca_{ij}=0\). If \(ca_{ij}=0\), then \(c\ne0\) permits division by \(c\), giving \(a_{ij}=0\). This proves the equivalence about zero entries. For a nonzero \(a_{ij}\), if \(c>0\), multiplication by \(c\) preserves whether the real number is positive or negative. If \(c<0\), multiplication by \(c\) reverses whether that real number is positive or negative. Therefore the asserted sign behavior holds at every nonzero entry. \(\square\)
These facts give useful checks, but they do not replace the definition. For instance, knowing that the zero entries stay in place does not tell us the numerical values of the other entries; those still have to be multiplied by \(c\). Likewise, knowing that a negative scalar reverses signs does not give the magnitudes. Calculate each product to obtain the full scalar multiple.
Common Mistakes and Corrections
Changing only some entries. A tempting but incorrect calculation is to multiply only the first row, or only entries that already have a minus sign. For example, take \(\mathbf{A}=\begin{bmatrix}1&-2\\3&0\end{bmatrix}\). The wrong calculation \(-2\mathbf{A}=\begin{bmatrix}-2&4\\3&0\end{bmatrix}\) leaves the entry \(3\) unchanged. The definition requires multiplying every entry, so the correct result is \(-2\mathbf{A}=\begin{bmatrix}-2&4\\-6&0\end{bmatrix}\). The correction is to write a product in every position before simplifying.
Changing the dimensions. Scalar multiplication does not add or remove rows or columns. If \(\mathbf{A}\) has \(m\) rows and \(n\) columns, then \(c\mathbf{A}\) also has \(m\) rows and \(n\) columns, including when \(c=0\). A zero scalar changes the entries to zero, not the shape of the array.
Misreading two negative signs. The product of a negative scalar and a negative entry is positive. For example, \((-3)(-4)=12\), not \(-12\). To avoid this error in a matrix, keep each entry product in parentheses until its sign is determined.
A complete scalar-multiplication calculation can therefore be checked in two passes: first verify that every output entry is the scalar times the input entry in the same position; then verify that the number of rows and columns has not changed. For a negative scalar, also check each sign. These checks apply to square and rectangular matrices alike.
Check Your Understanding
For each question, use the entry-by-entry definition and check the dimensions of your result.
- Compute \(4\begin{bmatrix}1&-2\\0&3\end{bmatrix}\).
- Find \(-\begin{bmatrix}-2&5&0\end{bmatrix}\).
- Compute \(\tfrac13\begin{bmatrix}3&-6\\9&0\end{bmatrix}\).
- What is \(0\mathbf{A}\) if \(\mathbf{A}\) has \(4\) rows and \(2\) columns? State both its entries and its dimensions.
- If \(c\ne0\) and \(c\mathbf{A}\) is a zero matrix, what must be true of \(\mathbf{A}\)? Give the theorem that justifies your answer.
