Autistic brains are more than messy desks: inching my way towards a study proposal

Learning new information is like filing papers.

Suppose you just got a new credit card bill from the bank.  You pay it, of course, but where do you put it in case you need to find it again?  (Maybe you need it to check whether you're reaching your target spending per month).  You probably have a folder for credit card statements.  If you have more than one credit card, you likely have a folder for that specific credit card.  You put the bill in the folder.  If you're really clever, you'll put it right in front so that the dates of the bills in the folder are arranged from front to back.  Next month, you want to check whether you're keeping within your budget, and realize you need your November statement.  You go straight to your credit card statement folder and pull out the paper in front.  Instant retrieval. 

Learning a new fact works similarly.  Suppose you've just learned that Neptune is a planet.  You go to your "planets" folder and add "Neptune" to it.  You also create a file for "Neptune," because now that you know it exists, you might learn other things about it.  Now as soon as you think of Neptune, you remember it's a planet, and every time you list the planets, Neptune is part of the list.

So when you learn, you group related ideas together based on something they have in common (like being planets, or being about Neptune), and you look under that commonality when you're ready to retrieve them.  When it comes to information, people call this process "organizing our knowledge" or "building a conceptual structure."

Many parents and teachers have been told that autistic people don't organize information this way.  Instead, their mental desk is supposed to look like this:

Above: a woman seated at a desk, dwarfed by a pile of papers that reaches the ceiling.  No neat stacks here--these papers face in every direction.  Another heap litters the floor around her desk.

Kathryn Stewart says that Aspergers and nonverbal learning disabled kids have "no internal organizational system--no hierarchy for ordering the information as more or less important."  As a result, "Information goes in and seems to get lost in a confusing internal filing system" where "everything is equally important" (from the 2002 edition of "Helping a Child with Nonverbal Learning Disorder or Asperger Syndrome"). Stewart isn't just some random blowhard; she directs the Orion Academy, a college-prep school for Aspergers and NVLD kids.

As far as I know, this claim has never been tested.  It probably comes from an outside point of view on ASD monologues.  Autistic people jump into a highly detailed discussion of their special interest without first orienting their listener to the important concepts.  One could conclude that autistic people know the important concepts, but don't explain them because they have difficulty figuring out what other people need to be told.  Instead, experts assume that autistic people don't have the important concepts at all. 

Yet, if you ask people on the spectrum about their special interests, you'll find they do have a conceptual structure.  If you ask, some can even tell you about it. 

For example, when Adam was about fourteen and interested in Dungeons and Dragons, he read the game's numerous several-hundred-page rule books cover to cover.  While he and his family were quietly doing separate activities in the same room, he sometimes would suddenly start talking about a (seemingly random) statistic from the game.  To his confused family, it sounded like number stew, but the information he shared contained patterns.  For example, your character has a certain amount of attack damage that goes up every time your character's level increases.  And characters with different jobs ("classes") increase in attack damage at different rates--a brawny fighter increases in attack damage faster than a slight wizard.  Adam could tell you what a fighter's or a wizard's "base attack damage" were at every level from 1 to 20.  He never spontaneously told his listeners that fighters and wizards scaled at different rates; what these rates were; that these were the extremes and other classes fit in the middle; or how this scaling matched the different classes' roles in the game.  But he could tell you all this, and more, when asked.  Where I, examining the same books, had failed to abstract the scaling rules from the sea of numbers in the tables, he had absorbed it.  Most likely, it was because he understood this meaningful structure that he could memorize so many vast tables of statistics.

Adam probably represents the rule rather than the exception.

I'm designing a study to demonstrate that a wide range of ASD adults--not just verbally gifted people like Adam--organize their knowledge.  To test this, I plan to interview autistic adults about their special interests.  They will be asked to write freely as much as they can on what they know about their special interest.   I will also interview neurotypical experts on the same topics (there's a vast literature on how they organize their knowledge).  I will then use a methodology called "concept mapping" to determine:

1. How do autistic adults organize their knowledge? (I have some predictions, based partly on Mottron and colleagues' veridical mapping theory).
2. Do autistic adults show as much organization as neurotypical experts? (I predict yes).
3. Do autistic adults organize information in a similar or different way to the experts? (This remains to be seen.  My guess is that they will be similar in some ways and different in others).

Above: a concept map that explains how a concept map works.

Designing a study like this requires resolving some difficult problems:
  1. What sort of open-ended questions will autistic adults interpret as an invitation to say whatever they want, rather than agonize over how to parse the question?
  2. How best to let autistic people know we're interested in their higher-level concepts as well as interesting examples, without distorting their answers?  I'm inclined to err on the side of being as explicit as possible, to avoid a problem found in certain studies about narrative.  Some experimenters tell autistic kids to talk about their lives, and do not explain that they are supposed to produce a narrative and that their responses will be judged according to a narrative model.  Not surprisingly, autistic kids answer the questions asked, but don't create a narrative--because they've been instructed to have a conversation, not to tell a story. However, in the same studies, they do produce narratives when explicitly asked.  I don't want to mislead autistic participants in this way.
  3. The experimental conditions likely to produce the best responses from autistic people (untimed writing; avoiding "open-ended" leading questions) might not produce the best responses from the neurotypical experts.  Is it better to make the procedure as similar as possible across ASD and NT participants, or is it better to make the length and quality of the responses as similar as possible, whatever that takes?

To solve these problems, looking at existing studies won't be enough, because they don't necessarily emphasize getting thorough responses from people with a wide range of verbal abilities.  Past studies that want participant input (e.g., studies of narrative), generally rely on spoken responses.  Thus, they exclude people without high expressive verbal ability.  I'd like to use written responses, and hope this will make it possible to include even people without communicative speech.

To do this, I need advice from the real experts: autistic adults themselves.  If you are an autistic adult and this project sounds interesting to you, please consider taking my survey. Click here to take survey 

I will be designing the study and drafting an overview (in the form of a mock grant proposal) over the next few months.  If I'm admitted to the appropriate graduate programs to work with specific professors, it could be conducted starting in fall 2014.   Your input could have practical impact within a year.
A FINAL NOTE: Look back at my examples of filing papers and Kathryn Stewart's quotation.  Did you notice that both involve a vertically-organized file system and think, "but that's not the only way to organize information?"  If so, we're thinking along similar lines. Autistic people do use vertical organization systems like categories, but probably rely less on them than neurotypicals (based on a line of research that has tried to find an autistic categorization deficit). However, I suspect ASD people use horizontal (same-level) relationships much more than neurotypicals, and their horizontal links are probably more sophisticated.


  1. Temple Grandin spoke of associative thinking, rather than linear, like the typical outline with roman numerals I learned in grade school. This is pictured here: http://commons.wikimedia.org/wiki/File:FreeMind-sample-mindmap.jpg There is a program one can download, but I hesitate to link to it, called freemind. (C-Net or download.com are not trustworthy to me, sourceforge may well be.)

    Strangely, in 1982, I learned about mind-mapping as a way to teach non-traditional learners to organize information, but I couldn't understand it. Not until I used it under advisement from an autistic adult for Ben, did I begin to understand.

    Is that what you are thinking? Am I even close?

    1. Wow, you're good! I'm actually thinking of using a more formalized version of mind mapping to analyze the data! (BTW, the software I'm using, CMap, is free and open to people involved in education, if you want to check it out).

      A concept map can be arranged in a lot of different ways, and you can ask a lot of different questions about how connected the concepts in it are, and in what ways. You can make a concept map of a very linear vertical organization system, for example. So using concept maps is just the first step. If you're interested, I can post some example maps.

      Interesting how mind-mapping was hard for you to understand at first. Was it easier for Ben? When I learned to use mind maps for data analysis in college, I realized I had been informally using them for years, and my mind basically works like one. At this point in my life many of the "nodes" are big concepts or theories with a link that may not be completely-specified. This lets me pack a ton of information into a small amount of mental space. Unfortunately, when I try to make what I believe to be a simple point, and start unpacking and justifying the link, I'm surprised to find myself writing several pages. (BTW, filing papers is difficult for me because a given article could easily belong in several different folders, so I have to either make copies or decide where I'd be most likely to want to reference it. Oddly, I've never seen this problem addressed in discussions about organizing one's desk).

  2. Mind mapping was a new idea, from what I remember. I have downloaded the cmap tool...I had to replace freemind. I'll learn how to use it, too.

    There are certain things I remember from 30 years ago that come back to me, although my degree was in ID, I took a couple of LD classes. (In Special Ed, it was ID ~or~ LD, two choices). Regarding LD, I remember being told when "these kids" get it, it comes as easily as breathing. They don't have to even think about it. All kids considered LD at that time had "special interests". I would love to see the comparison between the experts and the more auto-didactic learners --which is what John Elder Robison told us in a talk...they teach themselves. He spoke of the hours he spent learning his special interests and compared them to the hours a professional put in the regular educational environment. (How much time do college students spend learning, precisely?) He told the kids they could become an expert in that time they didn't spend socializing, in less time than a college student. I think you are on to something in this, and I would like to see it justified scientifically.

    Much luck!.

  3. It's strange that the writing by Stewart says the major categories for information are "important" and "unimportant" - and the other option is disorganization. That seems like one heck of a false dualism to me.

    Also, thank you for the next post you wrote - I was confused about what "horizontal" and "vertical" meant in this. The post helped me figure it out!