One perspective is extremely data-oriented and considers scientific research (ideally performed in collaboration with researchers who are themselves autistic) to be of paramount importance in characterizing autistic neurology and cognition.
As for why this characterization is important in the first place (aside from the science stuff frankly just being really interesting to a brain-geek like me), this perspective also figures that many of the problems faced by autistic people and our families stem from misconceptions about what autism is -- misconceptions which deny our strengths, misconstrue the reasons behind our behavior, and prompt all sorts of ridiculous assumptions about what we must intrinsically lack (e.g., self-insight, capacity for compassion, etc.).
This perspective is also the one from which I often feel compelled to point out, upon encountering people who invoke the standard litany of "will-nevers" ("she will never live in her own apartment or house, she will never make friends, she will never marry, she will never drive, she will never go to college", etc.) in response to learning that someone happens to be autistic, that these "will-nevers" are by no means justified by the facts.
Now, of course some people truly "will never" do one or more of the things on the above list; I'm presently in the "possibly will never drive" zone myself, for instance.
But nevertheless, it is a fact that autistic people (including some who had speech delays or other indicators often assumed to indicate "low functioning" as children) sometimes do end up growing up capable of doing many things nobody could have predicted when they were younger.
It is a fact that research has revealed areas of measurable strength common to many autistics.
And it is also a fact that, from an historical standpoint, the very first documented study of autistics (in which the word "autistic" was used), included the following observation:
Even though most of these children were at one time or another looked upon as feebleminded, they are all unquestioningly endowed with good cognitive potentialities...The astounding vocabulary of the speaking children, the excellent memory for events of several years before, the phenomenal rote memory for poems and names, and the precise recollection of complex patterns and sequences, bespeak good intelligence in the sense in which this word is commonly used.
- Leo Kanner, 1943
So, basically, from a scientific and observational standpoint, the notion that being autistic automatically brings with it a global cognitive deficit is just plain wrong. It was wrong in Kanner's time, and it is still wrong now. And the wrongness of claiming that someone has this or that deficit, in this context, has nothing to do with making value judgments about people and everything to do with perpetuating misconceptions that could lead to some people being essentially "written off" as not worth the time and effort it might take to educate them or provide them with enrichment opportunities.
The other perspective, however, is much more social-justice oriented, figuring that the longstanding disability rights movement (as well as the superset civil rights movement) have already done all the major philosophical work that autistic advocates need to draw on.
Additionally, this social justice perspective is in some ways orthogonal to the question of what autism is -- it is less concerned with trying to figure out how autistic brains work than with trying to figure out how to secure basic civil rights for people currently labeled or identified as autistic. It is even, at times, leery of fixating too much on the data the scientists like so much, as the data-gathering process (even when done well) might be backed by entities with a primary interest in, say, coming up with a means to identify autistic genes for the sole purpose of eugenically preventing future autistic people.
Moreover, there's also the matter of what the scientific perspective's focus on dispelling misconceptions leaves unsaid. If it is wrong to assume that all autistics have global, severe, cognitive deficits, does that make it somehow okay to write off people who do have global, severe cognitive deficits? I would personally say "absolutely not", and I don't think there's anything inherent in the scientific perspective that claims it's okay to abuse or neglect people once they can be considered "sufficiently disabled", but this kind of discussion does come up from time to time, and it does concern me.
A great example of something the social-justice sector of my brain wholeheartedly agrees with is the essay The Thing Itself Is The Abuse, from the Biodiverse Resistance blog.
It's not the inherent wrongness of the treatment that is discussed, it is the supposed "horrible mistake" of subjecting someone to that treatment when that person actually turned out to be not a member of the category of people that it's considered acceptable to do this sort of thing to. No thought is given to why it's supposedly "acceptable" to do it to people who are in that category, despite the fact that, in both cases, the reporting of the incident blatantly begs the question: if it was horrible and inhuman and inacceptable to do this to one person "by mistake", what is it to do it to a whole "Othered" class of people deliberately?
Clearly, there are at least two main "things" going on in all this -- first of all, there's the question of what it means for a person to be autistic (and the implications of this for advocacy, for who is considered "qualified" to self-advocate as an autistic person, etc.). And second of all, there are the very real (and extremely distressing) problems of abuse and neglect of, and failure to educate and provide opportunities to, individuals with all types of disabilities.
I want to be able to acknowledge and deal with both these things without having my efforts to do so come across as somehow being ignorant of the other -- e.g., if I point out that plenty of autistic people do actually go on to learn to type or speak, and make friends, I don't want this to come across as denigrating those who cannot type or speak even as adults. I care about them, too, and I would hope that anyone advocating for more ethical treatment of disabled persons would agree that it is not any particular "ability level" in anything that grants a person the "right not to be abused", but that the fundamental problem here lies in the assumption that certain kinds of people can and even should be subject to "extreme aversives" or levels of tranquilizers sufficient to stupefy them.
I have been trying for ages to figure out if these perspectives (the scientific and the social-justice) are compatible, and fundamentally I guess I believe that on some level they are, but it's so difficult to navigate the language around this stuff that I sometimes despair of ever being able to explain why and how my own overall perspective makes sense, even to me.
The closest I can get to that right now, I suppose, is saying that while social justice has nothing to fear from truth (as revealed by properly-done science), science is still a human endeavor, and therefore subject to untoward influence as far as how experiments are designed, conducted, and interpreted. And far from being "anti-scientific", autistic self-advocacy can directly support scientific endeavors by asserting ourselves as stakeholders, even as we work in other contexts toward better ethical standards for all people, regardless of presence, type, or level of disability.