Read for yourself:
DSM IV Criteria for Diagnosing Autistic Disorder* | Interactive Autism Network
DSM IV Criteria for Asperger's Disorder* | Interactive Autism Network
http://www.who.int/classifications/icd/en/GRNBOOK.pdf (ICD-10 criteria are in here)
In reality, a person could be diagnosed with Autistic Disorder in the DSM-IV or the ICD-10 without any speech/language delay and without any delay in self-help skills or intellectual/cognitive delays, as follows:
In the DSM-IV, one only needed to demonstrated a "
lack of varied, spontaneous, make-believe play or social imitative play appropriate to developmental level," as specified in the Communication category, in addition to the required number of symptoms in other categories.
In the ICD-10, the criteria are a bit more complicated to explain but a person could be diagnosed with "Childhood Autism" without any speech or language delay and without any self-help/cognitive delays, if they had "abnormal or impaired [...] functional or symbolic play" prior to age 3
[1], problems with social reciprocity/conversation and modulation/use of things like tone of voice and rhythm of speech, and at least 2/5 symptoms from the rigid/repetitive behavior section (which include obsessive/narrow interests and stimming/"repetitive motor movements", two things loads of aspies have) plus the usual exclusionary/"this must be the best explanation" criteria at the end.
You can go through the criteria for a classic autism diagnosis in either manual and create a symptom profile that meets the requirements for diagnosis of classic autism without the presence of speech/language delay or intellectual/cognitive delay.
On the flip side of things, lot of people who arguably meet criteria for classic autism (at least if you go strictly by the book) have been diagnosed with Asperger's or PDD-NOS instead.
Clinicians are not at all immune to the adoption and use of stereotypes when it comes to the way that they evaluate and diagnose people (they are only human -- fallible and sometimes even incompetent); Nor are they immune from choosing one diagnosis over another for pragmatic rather than scientific reasons, often reasons steeped heavily in social politics (diagnosis of Asperger's = less stigma and less risk of limiting a person's future social/vocational/academic opportunities, for example).....
Partly as a consequence of the aforementioned things; Partly because of how the diagnoses were written up
[2]; Partly because of a lack of knowledge about the etiology involved in either diagnosis and the heterogeneity of people with either diagnosis
[3]; Partly because of the many people not diagnosed until later childhood, adolescence or adulthood (when it may be difficult to confirm a person's early history) and the fact that people with HFA and Asperger's may have identifcal symptoms and neuropsychological profiles once they reach adulthood; And probably partly for other reasons I'm forgetting or unaware of,
there has never been a concensus among clinicians and/or researchers about where the boundaries lie between Asperger's and Autistic Disorder.
If you read in academic journals about the arguments for and against the separation of Asperger's Syndrome from other types of ASD, you will find loads of discussion about how clinicians did not adhere strictly to the criteria (or basically made up their own), about how there was quite a lot of disagreement in how the criteria should be interpreted and used, about whether the diagnostic criteria as written were or were not useful/valid.
There is evidence that what diagnosis a person with higher functioning autism received depended more on who diagnosed them than on what symptoms they did or did not display (in early childhood, at the time of diagnosis, or both).
----------------------------------------------
Footnotes:
1. Basically "make believe" or "social-imaginative" play like copying the social and practical behaviors of others and using toys in whatever way they were intended to be used. (As opposed to lining them up, sorting them by physical characteristics, repeatedly moving arms or spinning wheels, spinning them or throwing them, etc.)
2. In the DSM-IV, if a person met criteria for both classic autism and Asperger's, they were supposed to be diagnosed with classic autism (see criterion F) -- but this rule was often completely ignored in clinical practice.
In the ICD-10, it's entirely down to the clinician's judgement -- the relevent part of criterion E for Childhood Autism and criterion D for Asperger's Syndrome (the text is identical) states "The clinical picture is not attributable to the other varieties of pervasive developmental disorder".
3. Attempts have been made to find distinct neuropsychological profiles or to evalute particular areas of ability for the purpose of validating or invalidating a distinction between Asperger's and classic autism without intellectual disability. The results, across studies, are conflicted.
Even when a particular study does claim to provide support for having two separate conditions, the results are generally (perhaps always -- I can't claim to have read every study) useless because the logic is circular....all that anyone seems to have proven is that people with ASD can be separated into distinct categories based on particular areas of ability, not that there is any valid reason (e.g. etiological or therapeutic) to do so.
..........
Please note that making footnotes is not an attempt to be fancy or give the appearance that I'm some kind of expert (I'm absolutely not an expert, I have read a lot but I'm just an average layperson and I'm sure there are others on here who have read a lot more and have a lot more knowledge and understanding than I do) -- I just wanted to make my post slightly more readable, less tangential.