Andrew M. Simons, a professor of biology at Carleton University in Ottawa, Ontario, explains.
Like all "why" queries, the question of why men have nipples can be addressed on many levels. My four-year-old daughter, always suspicious of a trick when asked such obvious questions, answered: "because they grow them." In search of the trick answer, she quickly added that "chests would also look pretty funny with just hair."
Evolutionary biologists, whose job it is to explain variety in nature, are often expected to provide adaptive explanations for such "why" questions. Some traits may prove—through appropriate tests—to be best explained as adaptations; others have perfectly good evolutionary, but nonadaptive, explanations. This is because evolution is a process constrained by many factors including history, chance, and the mechanisms of heredity, which also explains why particular attributes of organisms are not as they would be had they been "designed" from scratch. Nipples in male mammals illustrate a constrained evolutionary result.
A human baby inherits one copy of every gene from his or her father and one copy of every gene from his or her mother. Inherited traits of a boy should thus be a combination of traits from both his parents. Thus, from a genetic perspective, the question should be turned around: How can males and females ever diverge if genes from both parents are inherited? We know that consistent differences between males and females (so-called sexual dimorphisms) are common--examples include bird plumage coloration and size dimorphism in insects. The only way such differences can evolve is if the same trait (color, for example) in males and females has become "uncoupled" at the genetic level. This happens if a trait is influenced by different genes in males and females, if it is under control of genes located on sex chromosomes, or if gene expression has evolved to be dependent on context (whether genes find themselves within a male or a female genome). The idea of the shared genetic basis of two traits (in this case in males and females) is known as a genetic correlation, and it is a quantity routinely measured by evolutionary geneticists. The evolutionary default is for males and females to share characters through genetic correlations.
The uncoupling of male and female traits occurs if there is selection for it: if the trait is important to the reproductive success of both males and females but the best or "optimal" trait is different for a male and a female. We would not expect such an uncoupling if the attribute is important in both sexes and the "optimal" value is similar in both sexes, nor would we expect uncoupling to evolve if the attribute is important to one sex but unimportant in the other. The latter is the case for nipples. Their advantage in females, in terms of reproductive success, is clear. But because the genetic "default" is for males and females to share characters, the presence of nipples in males is probably best explained as a genetic correlation that persists through lack of selection against them, rather than selection for them. Interestingly, though, it could be argued that the occurrence of problems associated with the male nipple, such as carcinoma, constitutes contemporary selection against them. In a sense, male nipples are analogous to vestigial structures such as the remnants of useless pelvic bones in whales: if they did much harm, they would have disappeared.
In a now-famous paper, Stephen Jay Gould and Richard C. Lewontin emphasize that we should not immediately assume that every trait has an adaptive explanation. Just as the spandrels of St. Mark's domed cathedral in Venice are simply an architectural consequence of the meeting of a vaulted ceiling with its supporting pillars, the presence of nipples in male mammals is a genetic architectural by-product of nipples in females. So, why do men have nipples? Because females do.