The verb 'subtract' entered English in the 1530s, modeled directly on Latin 'subtrahere' (past participle 'subtractus'), a compound of 'sub-' (from below, under) and 'trahere' (to draw, to pull). The literal Latin image is evocative: to draw something away from underneath, to pull away secretly or gradually. In Classical Latin, 'subtrahere' was used for stealthy removal — withdrawing troops, embezzling funds, stealing away.
The word's path into English was somewhat complicated. Before 'subtract' was adopted, English already had 'subduct' (from Latin 'subducere,' with a similar meaning) as its term for mathematical deduction. 'Subtract' gradually displaced 'subduct' during the sixteenth century, perhaps because the '-tract' form more transparently showed its connection to the productive 'trahere' family. The older 'subduct' survives
Mathematics gave 'subtract' its most prominent modern role. The four fundamental arithmetic operations — addition, subtraction, multiplication, and division — were codified in English mathematical writing during the sixteenth century. 'Subtraction' became the standard term for the operation of removing a quantity from another, and it remains the word every English-speaking child learns in elementary school. The mathematical sense is now so dominant that many speakers would struggle to use 'subtract' in any other context.
Yet the non-mathematical sense persists in formal English. One can say 'nothing can subtract from his achievement' (take away from) or 'the scandal subtracted from her reputation.' These uses preserve the Latin original's broader meaning of diminishment and removal, extending beyond mere arithmetic.
The Latin prefix 'sub-' (from below, under) adds a distinctive spatial dimension to the 'trahere' family. While 'extract' draws out, 'attract' draws toward, and 'distract' draws apart, 'subtract' draws from below — suggesting concealment, stealth, or undermining. This connotation appears in related Latin 'sub-' words: 'subterfuge' (literally 'fleeing underneath,' evasion), 'subterranean' (under the earth), 'subtle' (from 'subtilis,' finely woven, literally 'woven underneath').
In the history of mathematical notation, subtraction has been represented by various symbols. The minus sign (-) that we now take for granted was first used in print by Johannes Widmann in 1489. Before that, subtraction was indicated by words or abbreviations. The Latin word 'minus' (less), from which the minus sign takes its name, coexisted with 'subtract' in mathematical writing, and both terms remain in active use.
The word's phonology reflects a common English treatment of the Latin prefix 'sub-': the /b/ partially assimilates to the following consonant, producing the pronunciation /səbˈtɹækt/ rather than a fully articulated /sʌb/. The stress falls on the second syllable, following the standard pattern for Latin-derived verbs.
In computing, 'subtract' (or SUB) is a fundamental machine instruction present in every processor's instruction set. The arithmetic logic unit (ALU) performs subtraction as one of its basic operations, typically implementing it through the addition of a two's complement — meaning that at the hardware level, subtraction is actually a form of addition. This computational reality would have amused the Latin grammarians who saw 'subtrahere' and 'adtrahere' as opposite operations.