The AutSide Podcast

Cathy and I ask whose testimony counts as evidence. Autistic people describe the load, pain, and preparation behind observable behaviour—yet our accounts are too often dismissed before they enter the record.
Today, Cathy and I returned to a question that runs beneath so much of the argument about autism and gestalt processing: what counts as evidence, and who gets to decide? I can describe what happens inside my own experience—the accumulation of information, the difficulty recognising bodily signals, the labour of rebuilding a plan—and still find that description treated as something outside the evidentiary record. The behaviour is available for somebody else to interpret, whilst my account of what produced it becomes suspect. I am visible as an object of study. My standing as a witness is less secure.
I reached for the language of witness testimony because it makes that imbalance difficult to ignore. A witness’s account need not settle every question to belong in the record; it can be examined, compared with other accounts, and considered alongside other evidence. Autistic testimony deserves that same seriousness. My description cannot, by itself, establish a neurological mechanism or prove that every autistic person processes as I do. But those limits do not make it evidentially empty. There is a consequential difference between asking what an account can establish and deciding, before examining it, that the person giving it has nothing to contribute.
We explored that difference through my childhood difficulties with toileting. From outside, the problem might be framed as a failure to request permission or use the required phrase. From inside, there were questions about whether I could recognise the bodily signal at all, and whether language was accessible when I needed it. Requiring the correct words does not resolve either difficulty; it can make access to a basic bodily need depend upon the very capacity that is unavailable. My testimony brings those conditions into view, including the humiliation and trauma that a record of successful or unsuccessful toilet training might never capture. What an adult counted was not necessarily what the child endured.
Changing plans offered another example. I described how preparing for Saturday establishes a whole arrangement of expectations, language, bodily readiness, and practical decisions; moving the meeting to Sunday does not necessarily dissolve that arrangement. The first remains active whilst I construct the second. What looks like a small alteration from outside can become an accumulation from inside. Cathy recognised related accounts from other autistic adults, and we considered what happens when schools repeatedly change seating, schedules, activities, and meals, then explain a child’s eventual overwhelm through the last sound or disruption. The final event is easy to observe. The load that made it consequential requires a longer record—and attention to the person carrying it.
This also shaped our discussion of my developing paper on what “gestalt” can mean. I want a protocol that makes clear where a claim resides: in a description of cognitive architecture, a framework for understanding, or an intervention. Each raises different questions and requires evidence appropriate to those questions. A change in observable performance cannot automatically establish the architecture that produced it, just as testimony about an experience cannot automatically validate a particular intervention. Precision matters here because an argument against one claim should not quietly become permission to dismiss every other claim sharing its vocabulary—including the experiences people are trying to describe.
We ended with the distance between visible achievement and the support required to sustain it. People can see me teach, publish, and speak; they cannot necessarily see the preparation, unstable access to language, accessibility tools, pain, or narrow margin behind those outcomes. If competence becomes grounds for disbelieving my support needs, then the work of managing disability helps conceal the disability itself. That is why this conversation matters beyond a dispute over terminology. Autistic people are offering accounts of what support does, what demands cost, and what childhood experiences leave behind. Our testimony needs scrutiny, context, and careful interpretation. It also needs to be admitted.
The AutSide is a reader-supported publication. To support my work, consider becoming a paid subscriber.



This is a public episode. If you'd like to discuss this with other subscribers or get access to bonus episodes, visit autside.substack.com/subscribe

What does ‘gestalt’ mean across fields, and what evidence supports each use? This video introduces my developing paper’s floor-and-sense map. Subscribers are invited to examine the draft and help refine its distinctions.
An explanation can sound settled long before the argument it explains is ready. That is one reason I generate these video explainers during research: I want to hear what happens when my developing work is compressed into a sequence that promises to make it clear.
This video was generated in Google’s NotebookLM, using Gemini, from the rough draft of my upcoming paper, What ‘Gestalt’ Can Mean, and at Which Floor. It offers an early look at the work, and another layer of examination. An explainer has particular demands: it selects a route, establishes an order, decides what needs introducing, and moves the listener towards a conclusion. Listening to that sequence helps me examine whether the relationships I intended have survived.
Where does a distinction become an absolute? Where does a provisional map become a settled taxonomy? Where does the system supply certainty that my sources, or my argument, do not support?
This one is imperfect. It sometimes describes entire fields in sweeping terms, narrows perception to the visual, and presents the difficulty of a unifying definition as though its impossibility had been established. Those are places to return to the draft—to check whether my wording invited the interpretation, whether a boundary needs strengthening, or whether the system has added something of its own.
Its errors become questions for my work. They do not become my conclusions.
There are also quirks that are difficult to miss. It absolutely butchers the pronunciation of my name, finding several inventive ways to get it wrong. In my experience, names outside its familiar English naming patterns also require explicit pronoun instructions, or the system defaults to he/him. I specified she/her here; the narrator obligingly reads that instruction aloud in the introduction.
These details matter. A tool can help make someone’s work available whilst still failing to recognise the person whose work it is.
For me, as an autistic researcher with communication and language difficulties, this process provides another way to check what I have managed to communicate. I generate the research, the analysis, and the argument. The system returns an interpretation that I can listen to, question, and compare with what I meant. Understanding and language do not always become available together for me; hearing the work organised externally can help me locate what still needs expression.
That is part of what I want to make visible by sharing this video: how a tool can support participation, and how much active judgement that support requires.
The paper responds to a recent call for a clear definition of ‘gestalt’ by examining more than a century of its documented use. Across perceptual psychology, linguistics, developmental research, and communication practice, the term has described related forms of organisation without consistently describing the same object.
My purpose is to help stakeholders specify which sense they mean and what kind of claim they are making.
A floor identifies whether the claim concerns architecture, a developmental framework, or intervention. A sense identifies what is being described as whole within that claim. Those distinctions help us ask whether the evidence offered addresses the organisation proposed, and what further evidence would be needed to connect one level with another.
Programming several words onto one AAC button, for example, establishes an externally unified selection. How the person perceives, remembers, understands, or processes those words remains a further question. The relationship can be investigated once it has been stated clearly.
The paper proposes a specification standard to support that clarity. It leaves substantive questions about developmental stages, processor classification, and particular communication arrangements open, so that further research and practice can address them with greater precision.
I welcome readers into the draft whilst it is developing. Consider this an invitation to open, crowdsourced critical review: identify missing sources, question the distinctions, challenge an inference, or show me where the language carries more certainty than the evidence permits.
Where does the map help? Where does it need redrawing?
The video is one interpretation of the work. Your reading can help me strengthen the work itself. If you wish to participate, feel free to reply to the email copy of this post with your contact information.
The AutSide is a reader-supported publication. To support my work, consider becoming a paid subscriber.



This is a public episode. If you'd like to discuss this with other subscribers or get access to bonus episodes, visit autside.substack.com/subscribe

My developing paper traces over a century of ‘gestalt’ across fields, asking how its different senses can support a shared language for research and practice. Read the draft and help refine it through critical review.
Sometimes I need to hear my argument returned to me before I can hear where it has become vulnerable—where a distinction I thought I had established remains exposed to another reading, where a connection has been supplied rather than supported by the text, or where the language has not yet caught up with the work.
The audio shared here was generated by Google’s NotebookLM, using Gemini, from the outline of a paper I’m developing, What ‘Gestalt’ Can Mean, and at Which Floor. This is an early look at work in progress. The generated ‘hosts’ make sweeping claims that my outline does not warrant, and those errors are part of why I use this process.
I listen for what survives and what changes. Does the argument remain recognisable? Do its distinctions hold? Where does a carefully bounded proposition become a declaration about an entire field? When the system overreaches, I return to my own work and ask why—whether something in my wording invited that reading, whether an important boundary needs to become more explicit, or whether the system has supplied an argument of its own.
An error does not automatically identify a fault in my paper. It gives me a place to examine.
I am sharing this episode partly to offer a view into that recursive research process, and partly because the paper has implications across several fields. I also want to show how these tools can help someone with communication and language difficulties, like me, participate in the world. I generate the work—the research, the analysis, the argument. The system gives me another way to check what I have communicated, to hear what has become available to a reader and what remains insufficiently carried by the language.
That access matters. I have paid hundreds of dollars for an editor to examine an academic paper, only to find evidence that my work had been run through Claude and returned to me at the human editorial rate. In trying to find a human editor for my academic work, I encountered an economic problem as well as an accessibility problem: the human attention I was seeking had become difficult to obtain in a market where editors are expected to compete with Claude’s speed and cost.
Using the tool directly gives me quick and timely access to a checking process I can repeat, question, and inspect. For someone whose language does not reliably arrive at the same time as her understanding, that can help an argument reach the world.
The paper itself responds to a recent discussion of the need for a clear definition of ‘gestalt’. My response begins with the term’s documented history across more than a century—a history in which it has acquired several related, but distinct, uses. Perceptual organisation, linguistic structure, developmental direction, an inferred organisation within a learner, and an arrangement programmed into a communication system can all be described using this word. Each use asks something different of the evidence.
My purpose is to trace those uses and help stakeholders specify which one they mean. A single unifying definition risks concealing distinctions that research and practice need to preserve.
Two working terms organise the paper. A floor identifies the kind of claim being made: architecture, a developmental framework, or intervention. A sense identifies what ‘gestalt’ denotes within that claim—what, specifically, is being treated as a whole. Naming both helps us ask whether the evidence offered addresses the organisation claimed, and what further evidence would be needed to connect one level with another.
This is a contribution towards a shared language. The paper leaves substantive questions about developmental stages, processor classification, and particular communication arrangements open. Its examples examine the relationship between claims and evidence, including the language of my own earlier work. The generated hosts sometimes turn that examination into a critique broader than the one I am making. Their confidence should not be mistaken for my conclusion.
I am opening the draft to readers whilst it is still developing. Consider this an invitation to a form of crowdsourced peer review: read, comment, question the distinctions, identify omissions, and show me where a source or a claim needs closer attention. This is open critical reading, with responsibility for the final argument remaining mine.
Where does the paper help you understand what is being claimed? Where does it leave you uncertain? Where have I failed to preserve a distinction that matters?
Those are comments I welcome. They are part of how the work becomes more precise.
For subscribers who want to take part in the review, please reply to the email copy of this with your contact information.
The AutSide is a reader-supported publication. To support my work, consider becoming a paid subscriber.



This is a public episode. If you'd like to discuss this with other subscribers or get access to bonus episodes, visit autside.substack.com/subscribe

After weeks apart, Cathy and I returned to a question at the heart of my work: how does meaning become language? We followed its history across disciplines, brought it back to my classroom, and celebrated a sold-out webinar.
It was good to speak with Cathy again after several weeks apart. She returned with a deceptively simple question: what is gestalt, and where does a definition begin? That gave us a way into the work I’ve been doing whilst she was away—tracing a history that reaches well beyond Marge Blanc, Anne Peters, or Barry Prizant, and asking what has to happen for meaning to become language.
We talked about how easily that question gets divided among disciplines. Psychology, linguistics, speech and language therapy, and education each hold a part of it, but the connections can disappear when everyone works in a separate room. Much of the scholarship I’ve been reading approaches the question across those boundaries, often in languages that Anglophone discussions overlook. Cathy recognised that this is a global conversation, and it was lovely to be able to share where the trail through those papers has taken me.
The question matters in my classroom as well as in the literature. A child can learn that a kettle is hot before she can say, “The kettle is hot.” The experience has meaning; the architectural question is how that meaning becomes available in language. That same distinction runs through my current AutSide series: an observable behaviour may be called executive dysfunction, social deficit, or communication impairment, but the behaviour alone cannot tell us what produced it.
I also caught Cathy up on the conversation between my two Substacks. AutSide is where I’m working through questions of architecture, whilst Sensual Residue follows what it has been like to live with the consequences—including the life behind No Margin for Error. So often I do something first, then find the words to explain what happened and why. Writing lets me test whether the language carries the meaning I was trying to convey.
We ended with the news that Cathy and Ali’s webinar had filled all 400 places within 72 hours. I’m delighted to be taking part alongside others. More than anything, though, it was good to be back in conversation with Cathy—to pick up the thread, find that it was still there, and know we’ll speak again next week.
The AutSide is a reader-supported publication. To support my work, consider becoming a paid subscriber.



This is a public episode. If you'd like to discuss this with other subscribers or get access to bonus episodes, visit autside.substack.com/subscribe

Belonging is more than entering a room. In this video, I consider literacy, gestalt processing, visible strategy, and the structures that must change when every learner is fully understood as having belonged there all along.
This morning, I recorded a video about belonging from the ordinary room where much of my writing happens—in my home office and library, surrounded by books, papers, old technology, and the accumulated life of the work.
I chose flowers and pearls because I needed something lovely. My leg is not healing as it should, the pain remains significant, and getting dressed has become one of the ways I remind myself that pain does not get to possess the whole field. I wore red lipstick, too, and found myself smiling on camera—teeth crooked, tea-stained, chipped, and unmistakably lived in. I am beginning to like that smile. I am beginning, a wee bit more each day, to feel that I belong in this body.
That feeling belongs inside this video because its central argument is that inclusion and belonging are not the same.
Inclusion says: you may enter the room.
Belonging says: your presence tells us what this room must become.
I talk about students who enter literacy classrooms already carrying years of difficulty, observation, correction, and shame. They may be physically present, receive accommodations, and even become successful by institutional measures. None of that necessarily means they recognise themselves in the learning taking place.
I talk particularly about Self-Regulated Strategy Development as a way of making the hidden work of reading and writing visible. Accomplished readers and writers anticipate, hesitate, lose the thread, make connections, manage emotion, revise their understanding, and begin again. Yet schools often conceal this activity, show students only the finished product, and treat those who cannot infer the missing process as deficient.
For gestalt processors, meaning may arrive as a dense, simultaneous whole long before it can be presented in the sequence a classroom expects. We ask for the main idea before the dust has settled around the whole of the text—then mistake difficulty producing ordered language for an absence of organised thought.
Belonging begins when we stop treating the learner as a failed version of the person for whom the room was supposedly built. It begins when her presence changes our practice, her account of her own process carries authority, and difficulty does not revoke her membership.
Perhaps that is what I am learning in my own body, too.
Belonging is not the reward for becoming perfect.
It is the knowledge that I do not have to become someone else to remain.
The AutSide is a reader-supported publication. To support my work, consider becoming a paid subscriber.



This is a public episode. If you'd like to discuss this with other subscribers or get access to bonus episodes, visit autside.substack.com/subscribe