The term "people-first language" was first used in the disability community in 1988. However, it had its origins in the AIDS movement. In 1983, a group of activists took over the stage at the second National AIDS Forum in Denver, Colorado. They called for the right to participate in healthcare decisions, to be treated with dignity, and to be referred to as "people living with AIDS" not "AIDS victims" or other language they perceived to be derogatory. These demands became known as the Denver Principles and they changed healthcare forever.
The adoption of people first language was just one of the legacies of these principles. These principles also served to empower the AIDS community, encouraging them to start their own activist organizations and to fight for the resources and social support they needed. It was also the first time AIDS began to be defined as a social rights issue, with the speakers in Denver asking for equality under the law and protection from discrimination. (HIV/AIDS would later become protected under the Americans with Disabilities Act.) The Denver caucus humanized people living with AIDS by reminding listeners that, even with their disease, they are people first, and have the right to be loved and to love, to be cared for respectfully, and "to live and die with dignity."
We all understand why someone deserves to receive adequate and respectful medical care, but it is sometimes difficult to conceptualize why we must use specific language in the process. The easiest answer is, "because they ask us to do this," and we should all do whatever we can to respect the wishes of others. But there are deeper, albeit more abstracted, answers as well.
The principle of linguistic relativity, which has gone in and out of favor over the decades, essentially argues for strong ties between one's ability to articulate something and the way they perceive that something. The "strong" version of the principle states that language determines thought. However, a more tepid version of the principle has been developed and is more widely accepted; it says that words actually influence our thinking.
The most common, though inaccurate, example of this is the idea that "Eskimos have 100 words for snow." (This is untrue, and they're now known as Inuits, not Eskimos, but I digress.) Because they have 100 words, they see 100 different types of snow. We English speakers, meanwhile, have only one word (snow) and therefore can only see one thing (snow). If an Inuit asked you to go out and retrieve snow type #43, you would be incapable of doing so. In this way, our vocabulary, our language, our ability to articulate our surroundings, actually affects our perceptions.
Linguistic relativity, when applied to person-first language, would say that if we conceptualize of a person with paraplegia as a "paraplegic," we see them primarily as their disability and, in doing so, dehumanize them. If we use person-first language, however, we see them as a person first and foremost, who also happens to have paraplegia. In this way, we are sure to recognize the individual in question as a fellow human being and are less likely to dismiss them as "other" or something sub-human as a result of their physical differences.
This raises interesting questions when it comes to matters of racial and ethnical labeling. We call a black person a "black person" and the idea of calling them a "person who is black" seems strange. The same is true with a Pakistani, or any other nationality, though "a person from Pakistan" does roll of the tongue more easily than a "person who is black." Why person-first language has never been applied to minorities in this way I do not know, but if we are to take the principle of linguistic relativity at its word, and the progress of the disability community in our society as a testament to this way of speaking, then perhaps this is the next step towards equality: to disassociate human beings from their cultural heritage and the color of their skin.