Autism as Gradual Sensorimotor Difference: From Enactivism to Ethical Inclusion

Abstract

Autism research is increasingly moving to a view centred around sensorimotor atypicalities instead of traditional, ethically problematical, views predicated on social-cognitive deficits. We explore how an enactivist approach to autism illuminates how social differences, stereotypically associated with autism, arise from such sensorimotor atypicalities. Indeed, in a state space description, this can be taken as a skewing of sensorimotor variables that influences social interaction and so also enculturation and habituation. We argue that this construal leads to autism being treated on a par with other sensorimotor atypicalities such as blindness or atypical height. This leads to our conclusion that, insofar there is an ethical call to inclusion in our public sphere regardless of contingent bodily difference, an enactivist take on autism naturally leads to extending such inclusion to autism. Moreover, our analysis suggests a concrete way forward to achieve inclusion of autistics: by being more attentive to autistic sensorimotor specifics.

Full text

Penultimate draft. Accepted for publication in Topoi.
Please cite the published version.
1
Autism as gradual sensorimotor difference
From enactivism to ethical inclusion
Authors: Thomas van Es1*, Jo Bervoets2,3*
Affiliation:
1 Centre of Philosophical Psychology, University of Antwerp, Belgium
2 NeuroEpigenEthics, Department of Philosophy, University of Antwerp, Belgium
3 Leuven Autism Research Unit, Department of Health Sciences, Katholieke Universiteit
Leuven, Belgium
* Corresponding authors (both authors contributed equally to this project)
Abstract
Autism research is increasingly moving to a view centred around sensorimotor
atypicalities instead of traditional, ethically problematical, views predicated on
social-cognitive deficits. We explore how an enactivist approach to autism
illuminates how social differences, stereotypically associated with autism, arise
from such sensorimotor atypicalities. Indeed, in a state space description, this
can be taken as a skewing of sensorimotor variables that influences social
interaction and so also enculturation and habituation. We argue that this construal
leads to autism being treated on a par with other sensorimotor atypicalities such
as blindness or atypical height. This leads to our conclusion that, insofar there is
an ethical call to inclusion in our public sphere regardless of contingent bodily
difference, an enactivist take on autism naturally leads to extending such
inclusion to autism. Moreover, our analysis suggests a concrete way forward to
achieve inclusion of autistics: by being more attentive to autistic sensorimotor
specifics.
Key words: enactivism, autism, normativity, inclusion, sensorimotor, neurodiversity
1 Introduction
Autism has been intensively studied in the cognitive sciences. Initially, autism has therein been
primarily conceived of as a social-cognitive deficit. The most prominent suggestion is the (lack
of) Theory of Mind hypothesis (Baron-Cohen 2000), but other cognitive theories such as Weak
Central Coherence (Happé and Frith 2006) and (deficits of) Executive Functioning (Frith 1996)
are part of such deficit perspectives. Theories positing a social-cognitive deficit in autism have

Penultimate draft. Accepted for publication in Topoi.
Please cite the published version.
2
been criticized for unwarrantedly positing cognitive modules to perform representational
calculus in the brain (Hipolito et al. 2020), as well as for their ethical implications (McGeer
2007; Bervoets and Hens 2020). Severe criticism of autism as a deficit to social-cognitive
functioning also comes from the neurodiversity movement. Scholars like Milton (2017) and
Chapman (2020) argue that autism can be seen as a positive practical identity and should not
be seen as a flaw in autistic individuals1 but rather as a mismatch between autistic individuals
and their environment. In this context, recent autism research increasingly views autism as,
fundamentally, a sensorimotor atypicality. The ‘High and Inflexible Precision of Prediction
Error’ (HIPPEA) account (Van de Cruys et al. 2014) is a prominent example of this, but other
proposals in this vein are made by Mottron and colleagues (2006), by Markram and Markram
(2010) and by Murray (2018). All these theories are in the wide sense theories of sensorimotor
atypicality and therefore, as theories of a different autistic embodiment, amenable to enactive
treatment. Enactivism, broadly understood, is a strand in cognitive science that stresses the
importance of embodiment, autonomy and sensorimotor interactivity over representations and
brain states to understand cognition and behavior (Varela et al. 1991; Thompson 2007; Di Paolo
et al. 2017; Hutto and Myin 2017). These theories of autism are not wedded to a homogeneous
autistic phenotype consisting of social-cognitive deficits and instead naturally align with the
addition of ‘hypersensitivity to auditory, visual or tactile sensations’ in the DSM5 (APA 2013:
clause B4).2 The addition of hypersensitivity is in line with early characterizations of autism as
well as with autistic self-advocacy forming the initial impetus for the neurodiversity movement
(Pellicano 2013). Examples are violent reactions to sudden noise and fixation on specific visual
patterns.
Sensorimotor theories of autistic embodiment embrace the heterogeneity of the autistic
phenotype by trying to account in a statistical way for how social-cognitive issues may emerge
given sensorimotor atypicalities and societal and personal contexts. Constant et al., for instance,
(2020) develop such a notion of emergence based on HIPPEA. The crucial concept linking the
sensorimotor and societal contexts is scaffolding (Krueger and Maiese 2018).3 A scaffold, in
its literal meaning, is a structure of platforms typically used for the construction or renovation
1 In line with autistic preferences, cfr. Kenny et al. (2016), we use ‘identity-first terms’ throughout this paper along
related terms like ‘autism’ and ‘autistic’ that avoid one-to-one connection with terms like disorder or condition.
2 The DSM-5 also mentions hypo-responsivity but according to sensorimotor theories this can be seen as
connected to, and derivative of, hypersensitivity on sensorimotor accounts, see for instance Van de Cruys et al.
(2019).
3 Van de Cruys et al. (2014) and McGeer (2019) both independently also connect autism research with the notion
of scaffolding.

Penultimate draft. Accepted for publication in Topoi.
Please cite the published version.
3
of buildings, that allows construction workers to work at different heights. Used figuratively in
the interpersonal context, it refers to the practice of using material and social environmental
tools as the building blocks of novel forms of interaction: think of being taught, or of the digital
devices we use on a daily basis (Colombetti and Krueger 2015; Sterelny 2010). By focusing on
scaffolding, any specific trajectory of an autistic is never wholly reducible to autism as such,
but includes sensorimotor and social interactions totally specific to any autistic individual. The
sensorimotor, or perception/action-driven, approach of enactivism is more open to taking into
account autistic lived experience and - by that token - able to overcome the ethical issues, as
mentioned above, in an interdisciplinary and productive way (Bervoets and Hens 2020).
There is a lot of recent work (Hipólito et al 2020; van Grunsven 2020) addressing how
enactivism can lead to seeing autism in line with the neurodiversity tenets. That is, it does not
reduce issues associated with autism to an inherently problematic individual neurology.
Instead, these issues are seen as, at least in large part, the results of contingent social practices.
Think of shaking hands, making eye contact or meetings in noisy rooms. However, there is, as
of yet, no detailed bottom-up account of how specifically the enactive construal of autism as
the type of sensorimotor atypicality here considered entails an ethical and inclusionary
approach. Even if autism and enactivism have been linked in much depth via the key concept
of ‘participatory sense-making’ (Di Paolo & De Jaegher 2012; De Jaegher 2013; Fuchs & De
Jaegher 2009), a constructive account linking autistic embodiment via enactivist principles to
a better inclusion of neurodiversity has not been concretely given. In fact, an enactivist
treatment of autism and of inclusion of autistics and their lived experience is anything but self-
evident, see for instance van Grunsven (2020) who refers, inter alia, to Hutto (2003). Our
account shows instead how relaxing contingent social norms such as making eye contact can
concretely lead to including autistics by avoiding unnecessary stress and anxiety related to their
sensorimotor atypicality and this much in the same way as intercultural inclusion is promoted
by accepting diversity of greeting practices.
In this paper we shall explore how an autistic sensorimotor difference can lead to social
atypicalities that autism is stereotypically associated with. We will do this on the one hand by
employing a sensorimotor atypicality view on autism to show how skewing of certain variables
associated with sensorimotor interaction influences autistic (developmental) trajectories.4 We
will on the other hand rely on a critical reading of Linguistic Bodies (Di Paolo et al. 2018) to
4 Our treatment below relies on the HIPPEA account yet is not limited to this specific account. Its key premises
are limited to sensorimotor atypicality on the one hand and the enactivist framework on the other. This also means
(see section 4) that our argument is as such not limited to autism but extends to any sensorimotor difference.

Penultimate draft. Accepted for publication in Topoi.
Please cite the published version.
4
see how in such enactivist framework social and sensorimotor habits are mutually constrained,
leading to diverse tensions between autistic and neurotypical people. Taking both elements
together, we argue autism to be one of many gradually distributed sensorimotor differences
(such as height) which we overcome in each and every social encounter. Insofar the normative
ethical claim consists in furthering inclusion in our public sphere of people with contingent
bodily differences, an enactivist take on autism seamlessly leads to extending such inclusion to
autism. Given that our analysis is based on a precise understanding of autistic embodiment, it
also suggests a concrete way forward to achieve autistic inclusion through recognizing the
specific autistic sensorimotor atypicalities.
The paper is straightforwardly organized in three sections corresponding to the key
elements in the above paragraph. The next section deals with how autistic sensorimotor
atypicalities, as seen through the lens of HIPPEA, translate into a specific skewing of social
trajectories. Section 3 links this specific skewing to sensorimotor and social habit-forming, and
the effect this can have on intersubjective social autonomous interactions. In Section 4 we build
on this to develop our ethical claim. The argumentative structure is then as follows:
If, as we argue,
1) autism is fundamentally a gradual sensorimotor atypicality that, as
sensorimotor habits do, potentially branches out to social atypicalities, and,
2) broadly following Di Paolo et al. (2018), the overcoming of sensorimotor
differences is inherent to any social interaction,
then, we infer,
3) we should strive to include autistic ways of being and overcome stereotypes
as are typically associated with autism, furthering autistic inclusion in the public
sphere in the process.
That is, we first reiterate the HIPPEA literature that advocates for a sensorimotor definition of
autism. Our contribution here is to connect this with the enactive perspective and explore how
the sensorimotor perspective connects to the social one in the enactivist framework. Together,
we reach the statement (1). Statement (2) reiterates the enactivist view on how social interaction
presupposes an overcoming of sensorimotor difference. Our conclusion (3) connects the dots:
insofar social interaction presupposes overcoming sensorimotor difference, we ought to strive
for also overcoming autistic sensorimotor difference. What we do not argue for here is that the
relevant atypicality associated with autism is of a sensorimotor kind: this is already a position
in the literature that we build upon in HIPPEA for example. Similarly, we do not argue here
that a cognitive deficit view is incompatible with inclusion: this is already a position in the

Penultimate draft. Accepted for publication in Topoi.
Please cite the published version.
5
disability studies literature that we build upon, such as for example Barnes (2016). Furthermore,
we do not argue that the enactive perspective is necessarily better at reaching inclusion than a
cognitivist framework: this is not a comparative effort. Nor do we argue that enactivism
necessarily leads to inclusion as the literature suggests otherwise, see Van Grunsven (2020).
Given the structure of our argument, we suggest in our conclusion that this principled enactive
treatment of overcoming gradual differences - or differences in degree - can be leveraged to
also argue for better inclusion of people diagnosed with other cognitive or neurodevelopmental
‘disorders’.
2 Autism as atypical sensorimotor embodiment
The recent trend in autism research is to try to capture autistic atypicality as flowing from an
underlying sensorimotor difference. This aligns with common experiences reported in self-
advocacy and qualitative research focusing on sensory sensitivity (Hens & Langenberg 2018;
Pellicano 2013) and is already enshrined in the diagnostic criteria of DSM-5 (APA 2013).
Accepting atypical sensorimotor embodiment as a core commonality underlying autistic
heterogeneity allows a) to avoid ethical issues equating autism with higher cognitive
(dys)function (Bervoets and Hens 2020) and b) to link autism research to enactivist views that
– instead of relying on higher cognitive brain modules – rely on the notion of ‘scaffolding’
(Krueger and Maiese 2018). Scaffolding as a term in cognitive science has been popularized
by Sterelny (2010) and has been taken up in the enactive literature by Colombetti and Krueger
(2015). Essentially, it refers to the ways in which our material environment is used as a toolkit
that allows us to create novel forms of interaction. Metaphorically, one may visualize this as
using the metal pipes and wooden boards to create a platform to reach ‘higher’ places like actual
scaffolds in construction work. In reality, material tools are essential scaffolds in elaborating
our practices, but socially we also provide scaffolds for one another in teaching environments
(Vygotsky 1978).
Because of our approach, the benefits of our account are in principle compatible with
any sensorimotor difference postulated as underlying autism. However, we rely for the detailed
exposition of how sensorimotor atypicality concretely influences developmental trajectories on
the HIPPEA account (Van de Cruys et al. 2014). We do this because it allows us to make this
connection as precise as possible for the reader. Also, autistic scaffolding accounts based on
HIPPEA already have been proposed by, for instance, Constant and colleagues (2020), the
underlying Predictive Coding view has been theoretically linked to dynamic misattunement

Penultimate draft. Accepted for publication in Topoi.
Please cite the published version.
6
(Bolis et al. 2017) and such accounts are being intensively researched empirically, see Palmer
et al. (2015) or Zaidel et al. (2015).
In essence, what we want to show is that viewing autism as atypical sensorimotor
embodiment leads, in a state space description of the agent, to skewing certain variables
associated with the sensorimotor interaction that are crucial for social normativity (see section
3). Given such skewing we can then specify how autistic developmental trajectories are such
that they give rise to the type of social misattunement that has been so often mistaken for the
social-cognitive essence of autism5. Before we go into the specifics let us first illustrate this
with a familiar paradigm case of sensorimotor difference: blindness. Nobody would say that it
is impossible for blind people to get to certain states of social interaction. What is different
between blind and non-blind people is how they get to the physical place in which the social
interaction takes place. They clearly will not arrive there with the same environmental
(accessibility) conditions. The blind person will rely on, for example, a touching stick or
echolocation or acoustic guidance. The non-blind person will, typically, rely on visual cues and
would - failing those - feel quite lost. The sensorimotor difference then is generally accepted to
impact only the how of navigating physical space and not that it is possible to navigate to a
certain physical place. What we will argue below is that, in an enactive view of autism as a
sensorimotor atypicality, what is basically in question is the how of achieving certain states in
a physical (or more abstractly: state) space. Using blindness or deafness as an analogy
exemplifies how our approach aligns with that of the social model of disability where ‘minority
bodies’ can never be just seen as deficient bodies (Barnes 2016). Indeed, whilst there are clearly
obstacles to certain contingent ways of social interaction, such obstacles can always be met
(and, consequently, social interaction as such can always be achieved) in suitably adapting the
environment. One can take sign language interpreters or audio described movies as examples.6
HIPPEA, or in full “High and Inflexible Precision of Prediction Error in Autism”, is a
conception of autism within the predictive coding theory. Predictive coding in cognitive
science, roughly, is a theory on perception according to which the brain is essentially a
prediction machine that aims to minimize the divergence between its internally generated
prediction of the external world and the actual input (Hohwy 2013). In HIPPEA it is
5 In other words: we specify how this mistake stems from a “double empathy problem” (Milton 2017).
6 Two anonymous reviewers pointed out there might be specific social interactions inaccessible to some people
and we agree. Our ethical point is limited to the fact that what is considered essential to social wellbeing can be
made to become accessible regardless of sensorimotor atypicalities by assuming sufficient effort by all parties
involved. In fact, in line with our enactive treatment, we believe all individuals are to some extent particular from
sensorimotor point of view such that any successful social interaction presupposes overcoming such difference
(see section 3).

Penultimate draft. Accepted for publication in Topoi.
Please cite the published version.
7
hypothesized that autism consists in a specific way of dealing with perceptual uncertainty.
Following HIPPEA autism is an atypical way to process a difference (the “Prediction Error”)
between perceptual input and prior expectations. This atypicality then specifically consists in
atypically giving a “High and Inflexible” weight to the difference between both (in other words,
it consists in giving an atypical importance to “Precision”). Note that precision weighting is a
general feature of perception under Predictive Coding and that all individuals display variance
in both the weighting processes and the flexibility thereof. In other words: everyone engages in
precision weighting and displays a certain range of flexibility in this. HIPPEA thus by definition
denominates a difference in degree between non-autistic and autistic people.7
Predictive Coding is phrased as a perceptual theory but this - certainly on an enactivist
reading, see further below - implies action as well. Indeed, perception has been for a long time
recognized as an active process (Gibson 1986) of mutual engagement with one’s environment.
In this case of atypical importance of precision, we would expect to see an atypical sensorimotor
interaction with one’s environment; an interaction more attuned to the actual specifics of the
environment than to abstracting away from those specifics in order to flexibly switch from one
abstract form of interaction to another. This then leads to a learning style that is at odds with
the requirements of a typical environment. Given the aforementioned scaffolding, autistic
developmental trajectories will, but only statistically speaking (Bervoets and Hens 2020), lead
to emergence of social-cognitive elements of the DSM-5 (APA 2013). Let us make this more
precise with a specific example. When in a meeting, the autistic subject might be attuned to a
certain flicker of light or background noise, thereby giving less prominence to somebody new
entering in the room. At that time, the autistic subject might seem self-absorbed compared to
neurotypical others who are more rapidly cued to the new social situation, more quickly
meeting the new entrant by greeting them. This may at first sight seem just an innocuous
7 An anonymous reviewer pointed out that there may be different ethical implications of viewing autism as a
sensorimotor atypicality or as a deficit in cognitive function. As mentioned above we restrict ourselves in this
paper to the positive point related to ethical inclusion when seeing autism in an enactive way as sensorimotor
atypicality and do not pursue a negative point of seeing cognitive deficit as incompatible with ethical inclusion.
Relatedly, they pointed out that HIPPEA might be seen as a deficit in perceptual processing. Although we cannot
pursue this in the present paper, we call attention to the fact that, in predictive coding approaches, precision is a
general feature of all perception and that, therefore, differences in precision can only be seen as gradual differences
in line with the main argument in this paper (where HIPPEA is just used as an illustration). The inflexibility that
arises is, as we argue in Section 4, related to contingent social norms of switching between contexts, and it can,
therefore, not be interpreted as a contextless deficit within an individual. The upshot of this is that there can indeed
be no a priori bounds on inclusion based on such a gradual sensorimotor difference if social norms can never – in
line with what we argue below – be enforced rigidly. This does not mean that there will not be practical difficulties
or difficulties with respect to specific ways of social interaction. It just means, see section 4, that such tensions
can be overcome in allowing substantive social exchange.

Penultimate draft. Accepted for publication in Topoi.
Please cite the published version.
8
difference but in the context of learning it may be an element in a large series of missed
opportunities (or scaffolds) to attend to what is typically thought to be salient for development.
In fact, as worked out by Constant et al. (2020), the outcome of such an atypical learning
process may account for many atypicalities associated with the autistic self.
Let’s now take a closer, enactive look at how atypical precision may skew variables
which are associated with sensorimotor interaction. Specifically, in enactivism we should be
wary of reifying an atypicality related to precision as part of a computational module in the
brain. Instead, this atypicality is to be seen as capturing the specific way autistic bodies tend to
covary with their environment8. The difference between typical and atypical precision is a
difference in such covariation. Autistic bodies tend to covary in a more precise way with their
environments, picking up regularities in their environment in a more precise way. By that token
they will tend to be less sensitive to social habits which rely on abstracting from the concrete
perturbations in the environment (which, said in passing, can be seen as an enactive
reinterpretation of the terms “High” and “Inflexible” underlying the HIPPEA acronym).
Coming back to the analogy we made above with respect to blindness, this atypicality is not a
static feature of a ‘modeler’ or of ‘a model’ restricted to be in certain atypical states. Rather, it
concerns the more nuanced, atypical manner of attuning to one’s environment, i.e., an atypical
way of navigating states afforded by one’s environment.
Autistic atypicality is, then, maybe less ‘visible’ to external observers (noting that even
blindness may be compensated for by for instance echolocation and therefore remain rather
invisible, see Servick (2019)) but in itself very akin to classical sensorimotor differences such
as visual impairment. Autistic people will tend to rely a lot more on current tangibles in their
environment and much less on carrying out previously internalized habits aligned with social
conventions (for the enactive development of such a scaffolding, see: Krueger and Maiese
2018).9 In a physical space metaphor: they rather tend to “feel” their way around in the here
and now rather than “reenacting” behavior based on some abstract cues. This does not mean
8 For a complete argument to this effect see van Es (2020a; 2020b).
9 There is a bit of complexity here. DSM-5 (APA 2013) also specifies Restricted and Repetitive Behaviors. This
implies that autistic often display an insistence on sameness holding on to specific habits or repetitive patterns of
behavior. This might at first glance seem at odds with our reading of HIPPEA which is focused on the immediate
tangibility of the environment. However, in our view this can be seen as a developed coping strategy of autistics
in making their environment more predictable so as to be less overburdened by spurious changes. We cannot
develop this point further in the scope of this paper but see it as a crucial aspect of an enactivist reading of HIPPEA
in which such behaviors can be explained without reference to inflexibility of internal representations. We thank
Sander Van de Cruys for pointing this out to us. Whilst an important outstanding issue in autism research, we do
not believe its resolution is critical for the argument made in this paper. We also refer to footnote 3 on hyper- and
hyposensitivity for a similar connection between two apparently contradictory phenomena.

Penultimate draft. Accepted for publication in Topoi.
Please cite the published version.
9
they can get to less (or indeed more) places. It just means they tend to get there in a different
way (see also Hipólito and Hutto 2020). Take the example mentioned above: it is not because
the autistic subject seems self-absorbed that they mean any disrespect to the one entering the
meeting room. They are just more attentive to the ‘here and now’ of their sensory environment
and therefore prone to miss the sensorily less pronounced cues associated to what is typically
seen as socially salient. This will have no effect on the efficacy of the meeting as long as both
parties can agree this momentary desynchronization is of no consequence to the content of the
interaction.
This means that when we compare autistic and typical bodies, they navigate similar
territory or state spaces in different ways. An enactive reading of HIPPEA, and by extension of
the other differences in ‘sensorimotor apparatus’, leads to higher sensitivity to environmental
perturbation. Seen from a neurotypical eye, autistic embodiment is correlated to ‘overfitting’
sensory data but from an enactive point of view it just is another, atypical, way of navigating
one’s immediate environment. Given autistic bodies are more sensitive to (deal more
‘precisely’ with) their immediate environment, their actions depend more on the ‘tangible’ here
and now than on abstracting out of such tangibles based on ‘intangible’ cues that have become
associated with certain expectations of default interaction patterns. Phrased like this one can
see the link to other sensorimotor differences such as blindness: the difference with ‘typical’
agents lies in how certain states are reached and not in a principled restriction of what states
can, ultimately, be reached. This fits with the atypicality view of autism, opposed to the
deficiency view.
The skewing of how certain states are reached by an autistic body clearly has an impact
on sensorimotor interactions with its physical environment. It will be more sensitive to triggers
in such an environment, tending to covary with them in a more ‘precise’ way. This way it picks
up on regularities and patterns a ‘typical’ body will simply not pick up on. At this junction there
is neither value nor disvalue in this difference; it just is. However, if assessed by the standards
of habits formed by typical bodies or from the interaction patterns naturally sustained between
typical bodies, autistic differences will tend to be perceived as flouting the conventions, or the
norms, of ‘typical’ interaction and hence as some intrinsic ‘deficiency’.10 The relations between
10 That such a (dis)valuation of autistic difference as deficiency is compatible with enactive theorizing has been
shown by Janna van Grunsven (2020). The view from ‘typical bodies’ can be compatible with the enactive views
as developed by Gallagher (2004) and Hutto (2003) on autistic differences. An alternative enactive view based on
autonomous sustainability of interaction patterns is left open by the ‘participatory sense-making’ approach by Di
Paolo, Cuffari and De Jaegher (2018). In the next section we critically build on the latter approach in order to
capitalize on its strengths, at the same time avoiding this weakness.

Penultimate draft. Accepted for publication in Topoi.
Please cite the published version.
10
the individual differences, the emergence of shared norms and valuation are complex as can be
gauged from the above: an increased sensitivity to one’s physical environment may quickly be
perceived as an insensitivity to one’s social environment. In order to work out these relations
and their link to interactional differences between autistic and neurotypical people, we will in
the next section critically build on the enactivist framework of Linguistic Bodies (Di Paolo et
al. 2018) to see how social and sensorimotor habits are mutually constrained. This can then
serve as the springboard for our ethical argument to further the project of inclusion.
3 From sensorimotor atypicality to social normativity
In the enactive literature, Di Paolo et al. (2018) distinguish three distinct, yet intimately
intertwined levels of autonomous organization relevant to an organism: the biological and
sensorimotor levels of organization are tied to each individual, whereas the social level is
intersubjective, and unique to each social encounter. These levels are hierarchically layered so
that the biological level of organization lies at the foundation and enables the formation of any
other form of autonomy. Autonomy is defined as precarious operational closure (Di Paolo and
Thompson 2014; Maturana and Varela 1980). This means that an autonomous system consists
of a self-enabling process network, which captures operational closure, that is naturally
inclined to disintegration, which implies precariousness. A process network is self-enabling, if
each process in it enables or is enabled by at least one other process in the network. A process
enables another process when its continuation aids the continuation of the other process; and is
enabled by another process when the other process’s continuation aids its own continuation. In
this sense, the self-enabling nature of the network is what allows it to, albeit temporarily, resist
the natural incline towards disintegration.
Sensorimotor autonomy does not concern the physical integration of the system, but
instead its sensorimotor interactivity. Here too, it is thought, do we find clusters of activities or
processes that are self-enabling, and display a similar autonomous structure as we find in
biological autonomy. These autonomous sensorimotor activity clusters are also called habits
(Di Paolo et al. 2017; Barandiaran 2017). In the remainder of this paper, we will refer to these
clusters as interactional (sensorimotor) regularities. Crucially, sensorimotor autonomy is
irreducible to biological autonomy, and thus does not need to be in service of it either. Think
of the many routines that populate our daily activities, many of them irrelevant to biological
sustenance. Nonetheless, we do experience a sort of urge to maintain the activity clusters we
are used to, such as a need to see some light after having been in a dark room for an extended

Penultimate draft. Accepted for publication in Topoi.
Please cite the published version.
11
period of time. It is in this way that the habits that constitute an organism’s sensorimotor
autonomy can be thought of as autonomous, independent of an organism’s biological autonomy
(Di Paolo et al. 2017). Of course, the restriction of sensorimotor autonomy can have biological
effects; by way of stress, for example.
Social autonomy is intersubjective, and unique to each social encounter. This, following
Di Paolo et al. (2018), does not explicitly concern each individual’s autonomy, but instead the
autonomy of the social interaction itself. In this sense, each social interaction is characterized
as a self-enabling network (the ‘self’ here referring to the interaction, not to the ‘selves’ that
are interacting) that, unless actively maintained, naturally inclines towards disintegration. What
helps sustain an activity (and what does not) depends on the participants and the environment.
Arguably, not all three levels are actually associated with each organism. Some, the
argument goes, seem to only have biological autonomy so that their sensorimotor interactivity
can be reduced to, and is merely in service of, their biological autonomy, whereas others engage
at least in some sensorimotor activities for the sake of it: think of a pigeon rolling in the snow,
for example.
3.1 Levelled sensitivity
An autonomous system cannot be defined without at least implicitly defining the environment
it is embedded in. This is because in continuously enabling its own structural integrity – i.e.
maintaining its autonomy – a system necessarily interacts with its environment. This continuous
interaction can bias an autonomous system to the interactional regularities that are encountered.
Put differently, the interactional history of a system allows it to become attuned to the unfolding
interactions: to the particular ways that certain movements influence sensory signals (also
called sensorimotor contingencies, see O’Regan and Noë 2001; Di Paolo et al. 2017). A
sensitivity, then, is nothing more than a system’s structural bias towards certain interactional
patterns or regularities borne out of an interactional history, which includes biological,
sensorimotor and social interactivity. This structural bias could be unpacked in terms of reliable
covariance between the system and its environment (Bruineberg and Rietveld 2014; Hutto and
Myin 2017; van Es 2020).11 For two systems to covary, a change in one brings about a change
11 Di Paolo et al. (2018) and Di Paolo et al. (2017) take this a step further by taking a system’s precarious autonomy
to imply a normativity relative to each level of organization. More than sensitivity, this normativity is intended to
account for a supposed ‘judging,’ or ‘evaluating’ capacity. Each system’s biological autonomy thus implies an
‘inherent normativity’ which offers a standard of evaluation by which to judge the possibilities of interaction in
the environment as either harmful, positive, or neutral. There is a debate in the enactivist literature with regards to

Penultimate draft. Accepted for publication in Topoi.
Please cite the published version.
12
in the other. Reliable covariance is then a covariation relation that does not have a one-to-one
mapping but implies covariance to a certain degree. There is no definite threshold for this but,
in our case, minimally one may expect that the system’s structural integrity is to be maintained
for anything to count as reliable.
The sorts of interactional regularities that a system becomes sensitive to, depend of
course on the interactions it has been able to engage in: its interactional history. This means
that when considering sensitivities, we need to consider the system’s embodiment, which
determines the possible modes of interaction (Varela et al. 1991). After all, a system cannot
become sensitive to environmental changes that its senses cannot pick up on one way or
another. People, for example, cannot become sensitive to the ways particular winds affect one’s
flying capabilities, whereas birds are unlikely to become sensitive to a standardized pen grip.
Furthermore, interaction only happens in an environment. The sorts of interactions available
are thus also dependent on the environment one is in. Much like birds in a world that does
contain pens, humans only a few centuries ago could not have become attuned to pen-writing
activities either. Indeed, any activity is necessarily performed by a system, in an environment,
thus the possibilities for interactions are determined by the system and its environment, and so
the interactional history out of which interactional sensitivities are borne is, also, determined
by both the system and its historical trajectory through an environment.
A sensitivity is a system’s attunement to the particular way that certain movements
influence the system in one way or another. The different autonomous levels of organization
are influenced rather differently by one and the same motoric movement. Consider how,
biologically, eating food provides energy for continued sustenance, whereas sensorimotorically
the foods’ way of resisting the chewing pressure entails a certain texture, the influences on the
tongue as it swirls and moves the food about entail a certain flavor (Noë 2004). Socially, thus
intersubjectively, eating the food may prolong the interaction — not eating may be rude —
while simultaneously imply a potential break — perhaps the interaction is over once the food
is finished. The different ways particular interactions influence the respective autonomous
levels implies that the developed sensitivities are also different. Whereas biologically the
system may become attuned to, say, energy intake regularities such as the frequency or the
whether this individualistic account of normativity betrays a hidden representationalist commitment (Hutto and
Myin 2013, 2017). Our present argument does not hinge upon this debate. We can speak of interactional
sensitivities obtained on an individual level, and introduce normativity on a social, intersubjective level, not
perfectly in line with, but amenable to both takes in enactivism (Di Paolo et al. 2018; Di Paolo et al. 2017; Hutto
and Myin 2013, 2017).

Penultimate draft. Accepted for publication in Topoi.
Please cite the published version.
13
amounts, sensorimotorically the system may become attuned to textures and flavours, the
spaces food is eaten in, the tableware it is eaten out of or the tools it is eaten with. The individual
system may also become attuned to social regularities particular to food-eating practices, such
as the presence or absence of certain people, the appropriate manner of using the cutlery, and
whether or not to engage in, say, conversation.12
Though conceptually we can divide the three levels, in practice they are intimately
intertwined. We don’t, e.g., experience multiple levels of organization, nor do different levels
of embodiment refer to multiple physical bodies. There are however ways we can distinguish
them in practice. Consider that someone may sensorimotorically be a bespectacled person, so
that their spectacles are an inherent part of their ‘sensorimotor self’, whereas, biologically, the
spectacles do not figure in the self-enabling network. The same could arguably be said of
shoewear or even of clothes. Moreover, we sometimes find ourselves in conflict, when the
different interactional biases conflict; such as becoming peckish during an academic
presentation since, biologically, one may be attuned to energy intake at that time whereas,
socially, eating may be detrimental to a continued academic interaction. Conversely, each and
every activity is always associated with all three levels of autonomy, to the extent the respective
levels are associated with the system (Di Paolo et al. 2018). Every human activity minimally
takes energy, is some sort of sensorimotor interaction in a world that is inherently social in
nature. Though distinct, no level of autonomy ever ‘acts on its own’. It is in this particular
sense, that the interactional regularities that any system becomes attuned to are dependent on
its levels of autonomous embodiment as well as the environment it figures in.
3.2 How autism levels up
Now that the basic framework is laid out, we can start looking at how the sensorimotor
atypicalities common to autistic people fit in. Briefly rehearsing, the interactional regularities
one becomes sensitive to (interactional sensitivities) are dependent on a system’s interactional
history. Organismic interaction depends on organismic embodiment as well as environmental
12 There is a slight disparity here between the discussed levels of organization and their respective sensitivities.
According to Di Paolo et al. (2018), biological and sensorimotor autonomy are both individual, yet social
autonomy is inherently intersubjective. Following the in-text logic, this implies that the autonomy of the
interaction, instead of the participating individuals, becomes biased towards certain interactional patterns. There
may be something to this, as Di Paolo et al. (2018) show. Consider recurring fights with a friend that neither of
you actually wants to have. It may be, indeed, helpful to think of this in terms of a bias of the social encounter
itself, instead of one of the individual participants. However, in this paper, we shall also be concerned with the
individual interactional sensitivities pertaining to social encounters.

Penultimate draft. Accepted for publication in Topoi.
Please cite the published version.
14
constitution. This means that the interaction changes whenever the organism or the environment
changes; the organism comes to covary reliably with its environment.
The particular ways in which the organism has come to covary with its environment,
has become sensitive to its environment, has historically interacted with its environment, can
be described as a state space trajectory. Imagine a multi-dimensional state space, with one axis
for each variable associated with the system.13 The state of the system is then a multi-
dimensional coordinate, with a particular value for each variable. Whenever a variable changes,
the state of the system changes, pushing it to a different multi-dimensional coordinate. As an
organismic system is constantly in flux, over time we will see the changing coordinates of the
system as traversing a particular trajectory through the multi-dimensional state space.
Autism, we argue, is best taken to consist fundamentally in particular sensorimotor
atypicalities. Within the state space formalism, this implies that certain sensorimotor-related
variables in the state space may incline, or be skewed, to display particular values or
dynamics.14 This skewing is consistent throughout their developmental trajectory, and thus
consistently influences their modes of environmental interaction. Consequently, by impacting
interactions, it will over time impact interactional history, and thus interactional sensitivities.
Sensorimotorically, an autistic will thus display an atypical state space trajectory, meaning an
autistic becomes sensitive to different interactions and forms different habits. This may explain
certain common basic sensorimotor tendencies among autistics (Constant et al 2020). Yet it
underdetermines the exact influence it will have on any specific autistic individual. Each
trajectory is of course not only dependent on a particular skewing of certain variables, but it is
a holistic pathway: each and every aspect of the individual, as well as each and every aspect of
their environment affects the trajectory in their own ways. As such, not only is each trajectory
unique, but each autistic trajectory is also uniquely atypical. In this way, our model supports
the large heterogeneity in the atypicalities usually associated with autism. Now we can see how
this can ‘scale up’ to social interactivity (but see Gallagher 2017; Zahnoun 2019 for criticism
on the need to ‘scale up’).
13 What exactly it means for a variable to be ‘associated’ with a system is an interest-dependent matter. If we are
to model a person’s kinetic dynamics when going down a slide on a playground, variables pertaining to their
cycling habits are not considered ‘associated’ with the system for current purposes. However, when we are
interested in modeling this person’s risks for cardiovascular diseases, we may want to consider those. In this sense,
a state space description of a real-life system need not be complete to be scientifically valuable.
14 Whether the inclination is visible in particular variables or dynamics depends of course on which variables were
chosen to figure in the state space description. A system that accelerates strongly either skews to high values in
the ‘acceleration’ variable or quickly rising dynamics in the ‘movement speed’ variable. It depends on how we set
up the formalism.

Penultimate draft. Accepted for publication in Topoi.
Please cite the published version.
15
Per our explanation in Section 3.1, anything happening on the sensorimotor level of
embodiment impacts the system’s social interactional sensitivities. Sensorimotor atypicalities
entail a mode of interaction with the world different from that conventionalized by interaction
between typical people. As such, the social interactivity will be influenced by the autistic’s
sensorimotor atypicalities just as their picking up of objects is impacted by them. This starts at
the very fundament of social development. Consider that joint attention schemes, for example,
play out differently when one’s relevant sensorimotor habits are mutually attuned than when
they are not. Over time, this may result in atypical social habits, such as particular manners of
speaking, body language or a preference for certain environments to engage in social
interactions in. As such, consistent sensorimotor atypicalities will influence the social
interactional regularities that the agent becomes sensitive to. Put differently, sensorimotor
atypicalities entail atypical pattern-forming on a social level, which in itself implies a skewing
in the interactional regularities that the person becomes attuned to on a social level. Take
practices of joint attention - like looking each other in the eye - that are crucial for turn-taking
in the social realm. If, as Constant et al. (2020) argue, such practices are more difficult because
of sensorimotor atypicalities, they can lead to different trajectories with respect to social
preferences. At this juncture, it is important to understand that autism is just one example of a
possible skewing of a sensorimotor trajectory. Height, weight, or visual or acoustic acuity more
broadly may just as well be atypical and are accompanied by their own atypical habits both
sensorimotorically and socially.
As such, placing the atypicality of autistics at a sensorimotor level allows for the wide
variety of noted atypicalities in autistic social interaction (Jaswal and Akhtar 2019). One
person’s trajectory may have resulted in habits that eschew fantastical imaginings (see, for
example, DSM-5 (APA 2013) or Curry and Ravencroft (2002)), whereas another’s interactional
history may have resulted in habitual interaction in fantastical worlds (see, for example, Lyons
and Fitzgerald (2012)). However, they may, coincidentally, share a habit of avoiding eye
contact (Valiyamattam et al. 2020). They also may not (Jones et al. 2017). This all depends on
the particular state space trajectory of each individual, as it does for every person, autistic or
not. Moreover, this trajectory is a sum of the organism’s state space and its environment’s,
which means that the same person will develop different habits if it were to have grown up in
a different environment. This is to say that from that point on, their state space trajectories will
come to vary wildly, and they will grow up to be different persons altogether.

Penultimate draft. Accepted for publication in Topoi.
Please cite the published version.
16
3.3 From individual sensitivity to social normativity
Individually, any organism has its own embodiment accompanied by its own set of sensitivities
on a biological, a sensorimotor and a social level. However, according to Di Paolo et al. (2018),
each social interaction also has its own autonomy. Following the definition of autonomy above,
this implies that each social interaction is composed of a self-enabling network that naturally
inclines towards disintegration. This means that, if the processes in the network were to stop,
the entire interaction would break. Only by continued processual activity can the whole network
be kept active: a conversation cannot be sustained without continued interactivity.
Moreover, each social encounter is unique and entails its own state space description.
The potential modes of interaction that could either sustain or break the intersubjective
autonomous organization are thus also unique to each social encounter. Like the organismic
case, what sustains or breaks a social interaction is dependent on the processes that make up
the network itself, as well as the environment it is embedded in. Naturally there is some
variability in participants’ sensitivities: some people may be used to speaking at a particular
volume that may seem too loud or too soft to others. Yet some environments also may afford
speaking loudly better than others. The range of speaking volume which best sustains or breaks
a particular social interaction is thus dependent on both the participants’ sensitivities as well as
the environment the interaction figures in. Indeed, contrary to the individual organismic case,
the interactional processes that make up this network are composed of those of all participants,
and thus depend, but cannot be reduced to, all the individual participants’ own interactional
sensitivities (Di Paolo et al. 2018).
This makes the interactional state space description an intersection of the individual
interactional regularities and sensitivities of at least two participants, their environment, and
their individual relation to the shared environment. An atypically tall person for example may
well have adjusted to looking slightly downward in social interactions, which means they may
not be well attuned to social interactions with someone even taller. As it is for the atypically
tall person, so it is for the autistic. This is to say that an autistic’s own interactional history will
in part determine what will sustain and what will break a social encounter they are engaged in,
just as much as their interactional partner’s interactional history does. This illustrates how a
gradual sensorimotor difference can over time reverberate into a qualitatively different social
experience, dependent on the interplay of how pronounced such difference is as well as on how
much one’s environment is willing to accommodate such a difference. Whilst tallness is not
immediately seen as consequential, consider shortness: one may not reach certain items in a

Penultimate draft. Accepted for publication in Topoi.
Please cite the published version.
17
supermarket, may not be able to use public restroom facilities, or, indeed, may not be able to
look interlocutors in the eye. Indeed, although this concerns a difference in degree in one and
the same quantitative dimension, it is immediately apparent how with current social norms this
will translate easily into a disabling condition affecting social interaction in a lasting way.
Furthermore, the analogy between atypical length and the case of autism with respect
to eye contact can be taken literally. In both cases there are sensorimotor reasons complicating
eye contact: differential tallness in the one case, and hypersensitivity to the other’s gaze in the
other. Yet whether or not this develops into a lasting preference to avoid eye contact depends
to a large extent on the reactions of others for whom there are no such complicating factors. If
they insist on it, this may well make any meaningful social interaction problematic. If this
problem is one encountered by individuals from very early on in development — as seems to
be the case for many autistic people — then this may result in a lasting tension with respect to
making eye contact. This tension in the social encounter needs to be overcome if the social
encounter is to be sustained, but it is not different in kind to any other tension specific to each
and every social encounter between two uniquely different individuals embedded in a unique
environment (Di Paolo et al. 2018). In fact, when the environment actively makes the difference
inconsequential, it is less likely to become such a focus of lasting, problematic tension.
It is this theme of overcoming tensions and differences that we take up in the next
section where we link it to the ethical theme of inclusivity of diversity. What we have argued
for in section 2 and 3 respectively is that 1) autism is fundamentally a gradual sensorimotor
atypicality which branches out to social atypicalities, and that 2) overcoming of sensorimotor
differences is, in general, inherent to any social interaction. Section 4 then infers that we should,
on the same grounds as accepted for promoting the inclusion of other diversities, also overcome
the social tensions flowing from sensorimotor differences associated with neurodiversity.
4 The ethical value of overcoming differences in degree
It is hardly controversial that differences in visual ability or height should not lead to exclusion
of those on the extremes of our human spectrum from our community. However, we can also
arrive at a principle of inclusion via classical ethical literature. Indeed, grounding of our moral
stance in our social communicative practices is, for instance, given by Strawson (2008/1962)

Penultimate draft. Accepted for publication in Topoi.
Please cite the published version.
18
via his notion of ‘reactive attitudes’.15 Crucial in this leading ethical framework is negotiation
of ethical interaction across individuals characterized by individual differences. Indeed,
Wallace (2019) develops a call to cosmopolitanism using reactive attitudes where we should
maximally broaden our ethical community irrespective of contingent social and cultural
practices and merely based on the possibility of entering into a mutuality of reactive attitudes.
In another Strawsonian interpretation, McGeer (2019) works out an inclusionary
account based on the ability to scaffold towards social communicative practices. She contrasts
this to approaches where certain specific social capacities are required as preconditions of being
full members of the moral community. Her approach avoids setting minimal entry conditions
based on contingent social practices such as eye contact as prerequisite for reactive attitudes.
Instead, she focuses, in line with our above argument, on the ability to scaffold regardless of
specific initial sensorimotor differences. Obviously, this requires both parties to the interaction
to put in an effort in order to overcome such differences. This is also in line with disability
literature in general where inclusion is central, and this independent of meta-ethical debates.
This is clear from, for instance, Barnes’ take on The Minority Body (2016), and neurodiversity
literature specifically. As a particularly acute example we can take Milton’s (2017) expression
of the double empathy problem. In the latter, Milton argues that understanding autism as a
deficit in social-cognitive ability actually is tantamount to a lack of empathy from the point of
view of a neurotypical agent. In the former, Barnes argues that attributing issues to individual
disabled bodies is, equivalently, an inability to see from the perspective of typical, able, bodies
how those issues actually stem, at least for the largest part, from a lack of societal concern for
accessibility of minority bodies. Our argument can be seen as integrating these into a Minority
Brain view (Bervoets and Hens 2020). Indeed, overcoming visible sensorimotor differences
requires the non-disabled others to put in effort to overcome related accessibility issues. In the
same way we argue that prima facie invisible neurological differences give rise to an analogical
15 The ‘reactive attitudes’ are attitudes of blame, resentment, forgiveness, gratitude etc. which we discern in the
other and the other in us. They allow us to build practices in which we can hold each other responsible. There are
two remarks to be made about our using this framework as an illustration of our argument. First, there is a lot of
controversy on whether such an account does not risk excluding others based on the ability to recognize and show
the reactive attitudes. It has been used as a way to rationalize either (partial) exclusion of autistics (Shoemaker
2015) or to positively argue for a ‘normalization’ of autistics to recognize conventional social cues (Richman and
Bidshahri 2018). However, we use this theory for illustration precisely because of its ability to abstract away from
sensorimotor (a)typicalities and its consequent bringing into focus of a communicative interaction as such. Second,
we use this theory, just like we used HIPPEA above, as a concrete illustration of our argument without it being a
necessary premise in what we contend. Any ethical framework that can develop a need for inclusivity over and
above individual differences will do to support our conclusion.

Penultimate draft. Accepted for publication in Topoi.
Please cite the published version.
19
societal responsibility: adapting neurotypical social practices in order to allow for better social
accessibility to the neurodivergent (autism being our prototypical example of neurodiversity).
Now, as we have seen in section 3, the sustenance of social interactions is in a crucial
way independent of specific underlying sensorimotor preferences. We do not have to react in a
specific way to be able to signal, for instance, blame or gratitude. Exactly like blind and seeing
people navigate (see our discussion in section 2) the same physical space differently, tall people
and people of average height navigate differently the interactional state space so as to recognize
reactions of blame or gratitude. The ethical call to inclusion is to separate the how we achieve
an ethical interaction from the that, the very possibility, of reaching such interactions. Here we
need to clearly separate the unique trajectories individuals traverse in order to reach some point
of social and ethical interaction and the fact that such points can be reached. In enactive terms:
whilst everybody lays down a different path while walking (Thompson 2007), there is no reason
why our paths may not intersect in participatory sense-making (Di Paolo et al 2018, specifically
the example worked out with respect to autism in Chapter 8).
Let us make this more precise with respect to autism and autistic embodiment. Where
autistic people are sensorimotorically skewed to attune more precisely to the here and now of
environmental perturbations, this is not the case for the majority of (neuro)typical people. It is
in this way understandable that the latter mistake their specific way of coming to reactive
attitudes for minimal, or essential, preconditions of achieving them. Whether it is by shaking
hands, recognizing specific facial expressions or in other ways abstracting from environmental
cues deemed inessential, typical people will tend to mistake a sensorimotor preference for
certain habits in social interaction for the essence of that social interaction. In the above we
have given the example of eye contact and other contingent social practices that may be self-
evident for neurotypical people but, for sensorimotor reasons, hard for those deviating from the
statistical sensorimotor norm. That these practices are contingent to the social interaction itself
can easily be seen from the possibility to interact socially without eye contact, and they should
thus not be seen as essential barriers to social entry. Autistic people will nevertheless from
childhood on have difficulty attuning to such habits precisely because of a statistical atypicality
in the way in which they tend to covary with environmental cues. We have shown that this
skewing of the variables associated with navigating the multi-dimensional state space should
not be reified into stereotypes of social interaction, because it is but one of a great many
historical factors that contribute to how unique autistic individuals wind up trying to engage in
social interactions optimized for a sensorimotor embodiment that is not theirs. In a society more
forgiving of the preconditions of expressing reactive attitudes, they may more easily develop

Penultimate draft. Accepted for publication in Topoi.
Please cite the published version.
20
shared habits than in one that is less forgiving. Obviously also, the longer it takes to be
recognized as one who is able to communicate, the more this part of one’s history will further
skew one’s potential to bridge the gap between one’s own sensorimotor preferences and the
social interaction which is deemed foundational for inclusion in our moral community. A
specifically acute example of overcoming such gaps is given in the context of early language
development by De Jaegher (Di Paolo et al. 2018: Chapter 8).
None of this is to say that these social, ethical, interactions are disembodied. Far from
it, in line with Di Paolo et al. (2018), each human interaction is necessarily embodied. But,
because of the relative autonomy of sensorimotor and social interactions we are not limited to
a specific way, a specific habit, of sensorimotor interaction to achieve a specific, ethical
interaction. Rather, by prescribing specific sensorimotor interactions as basic for ethical
inclusion we violate the independence of the social and ethical field and – see our argument in
the previous paragraph – risk to actively contribute to exclusion through ‘ableist’ bias. Indeed,
there is a sense in which all of disability, at least the social model aspects of disability, can be
seen as the result of too prescriptive a notion of sensorimotor interaction. Such too prescriptive
notions literally exclude people from ethical interaction by making such interaction inaccessible
for those who are sensorimotorically different.
As is clear from the above, the overcoming of differences in degree (whether in visual
acuity, height, culture, ...) is something we do daily among people. We recognize just societies
by their capacity to overcome the superficial tensions that may arise from such differences, i.e.,
their tendency to embrace cosmopolitanism, per Wallace (2019). What we argued for is that
the enactive construal of autism as an atypical embodiment leading to a specific skewing of
sensorimotor preferences leads to recognizing it as one such difference in degree. Looking at
how such difference is overcome in actual lived practice helps us explain the precise nature of
this underlying difference and - consequently - may help us to create more inclusive conditions
by relaxing sensorimotor requirements which underlie social interaction of (neuro)typical
people. We have mostly given relatively straightforward examples which might seem relatively
easy to overcome, such as whether to maintain eye contact. However, our argument (as per fn.
7) does not admit of a priori bounds on conditions to be relaxed. To give a concrete example
of this: think of non-verbal autistics. One may be tempted to consider verbality a minimal
condition for social interaction, but this is challenged, amongst others by disability scholars in
an intriguing case study (Van Goidsenhoven and De Schauwer 2020). In fact, the practice of
living with autistic people shows that what was once seen as an insurmountable barrier can be

Penultimate draft. Accepted for publication in Topoi.
Please cite the published version.
21
overcome and scaffolding to rich communication can be achieved in a variety of ways, see also
De Jaegher (2013).
Clearly, there is a cost to cosmopolitanism on the side of typical people who need to
adapt to new habits (or unlearn old habits), but we believe that not doing so (and consequently
leaving all the cost of adaptation to the atypical) is a case of bigotry. This does not mean that
the final result will not be a compromise between a cost incurred on both sides - see the double
empathy problem of Milton (2017) - but it does mean there is an ethical (as well as scientific)
call to at least figure out how to come to a better compromise, even if this is a path we can only
lay down in walking.
5 Conclusion
In conclusion, it may be relevant to revisit our argumentative structure. By connecting HIPPEA
with enactivism, we have argued that (1) autism is fundamentally a gradual sensorimotor
atypicality that, as sensorimotor habits do, potentially branches out to social atypicalities. Then,
(2) broadly following Di Paolo et al. (2018), we argued that the overcoming of sensorimotor
differences is inherent to any social interaction. Indeed, everyone’s interactional regularities,
or their developmental trajectory more broadly, influences their concrete social interactions.
Differences in the way one approaches social interactions can cause tensions to be overcome
in any social interaction. According to our analysis of autism, it is just one of the many ways
in which one can be sensorimotorically atypical. As such, we conclude with (3) we should strive
to include autistic ways of being, furthering in the process more inclusivity in the public sphere
(McGeer 2019).
We have argued that our enactivist construal of autism entails an inclusionary approach
in line with the neurodiversity tenets specifically and cosmopolitarian ethics generally.16 That
this is not self-evident is clear from other enactivist approaches to autistic embodiment, see
Van Grunsven (2020), which start either from the neurotypical interactional habits (Hutto 2003;
Gallagher 2003) or from the mere autonomy of specific sensorimotor interactions (Di Paolo et
al. 2018). Our approach instead starts from a relative - and ethical - independence of social
interaction and so leaves room for an inclusionary approach to autistic embodiment seen in its
sensorimotor specificity. Basically, we show that there are many ways to peel an orange and in
our daily practice we overcome sensorimotor differences that unavoidably arise from different
16 This does not imply that a cognitivist construal cannot arrive at inclusion in a different way.

Penultimate draft. Accepted for publication in Topoi.
Please cite the published version.
22
life histories. Autistic embodiment is just one of such bodily differences. Understanding autistic
sensorimotor difference through a good empirical theory of autism, such as atypical precision,
allows us to direct our search for practical ways in which we can increase mutual understanding.
After all, we are all continuously changing bodies in need of being cut slack in order to maintain
social, ethical, interaction.
In the process, we have also engaged with the enactivist literature. We have expanded
enactivism to provide a space for speaking of individual social sensitivities, over and above the
biological and sensorimotor individual sensitivities. This has allowed us to expand the merely
descriptive enactivist analysis into the normative debate. This provides new tools for the ethical
project of inclusion, by specifying particular social tensions that we should work to overcome,
as we do on a daily basis elsewhere. With this, we avoid slipping into enactivist developments
of autistic embodiment that, ultimately, point back to basic, or essential, social capabilities and
therefore deficiencies related to autism (Van Grunsven 2020). Furthermore, we connected the
enactivist notion of scaffolding with its use in McGeer’s work, allowing for an ethical shift of
focus from the overt capacities typically used in social interactivity to the minimally - shared -
ability allowing social interactivity at all (McGeer 2019; Krueger and Maiese 2018).
Crucially, the toolkit provided here can be put to use in other disability studies as well,
such as Tourette Syndrome or ADHD, and may serve as a general framework from which to
analyse and help resolve social tensions. In general, this can bring ethics and empirical science
more closely together via enactivism. However, we do not think a philosophical framework is
the solution to inclusion. If this were the case the struggle for better inclusion of ‘physical’
disabilities would long be over. The solution lies in individuals ‘actively’ trying to overcome
difference, thereby providing the evidence that this overcoming is possible and, hence, that the
difference is not a difference in human quality but merely one of physical degree. Still, this
framework can be, in our opinion, a good support for such activism in showing that inclusion
should not be a mere targeting of individual difference but requires an effort on both sides of
the interaction.
Funding
This work is partly funded (JB) by the NeuroEpigenEthics project that received funding from
the European Research Council (ERC) under the European Union’s Horizon 2020 research and
innovation program (grant agreement 804881), and (TvE) by the Research Foundation Flanders
(FWO), grant number 1124818N.

Penultimate draft. Accepted for publication in Topoi.
Please cite the published version.
23
Acknowledgements
We want to acknowledge the invaluable comments of Kristien Hens, Inês Hipólito, Sander Van
de Cruys and Damian Milton for helping formulate our argument as presented in this paper.
Bibliography
American Psychiatric Association (2013). Diagnostic and Statistical Manual of Mental
Disorders (DSM-5§). Washington, DC: American Psychiatric Association Publishing. doi:
10.1176/appi.books.97808904 25596
Barandiaran, X. E. (2017). Autonomy and enactivism: towards a theory of sensorimotor
autonomous agency. Topoi 36, 409–430. doi:10.1007/ s11245- 016- 9365- 4.
Barnes, Elizabeth. (2016). The Minority Body: A Theory of Disability. Oxford
University Press.
Baron-Cohen, S. (2000). Theory of mind and autism: a review. Int. Rev. Res. Ment.
Retard. 23, 169–184. doi: 10.1016/s0074-7750(00)80010-5
Bervoets, Jo, and Kristien Hens. (2020). Going Beyond the Catch-22 of Autism
Diagnosis and Research. The Moral Implications of (Not) Asking ‘What Is Autism? Frontiers
in Psychology. https://doi.org/10.3389/fpsyg.2020.529193.
Bolis, D., Balsters, J., Wenderoth, N., Becchio, C., and Schilbach, L. (2017). Beyond
autism: introducing the dialectical misattunement hypothesis and a Bayesian account of
intersubjectivity. Psychopathology 50, 355–372. doi: 10. 1159/000484353
Bruineberg, J., & Rietveld, E. (2014). Self-organization, free energy minimization,
and optimal grip on a field of affordances. Frontiers in Human Neuroscience, 8, Article 599.
Chapman, R. (2020). The reality of autism: on the metaphysics of disorder and
diversity. Philos. Psychol. 33, 799–819. doi: 10.1080/09515089.2020.17 51103
Colombetti, G. & Krueger, J. (2015) Scaffoldings of the affective mind, Philosophical
Psychology, 28:8, 1157-1176, DOI: 10.1080/09515089.2014.976334
Constant, A., Bervoets, J. Hens, K., and Van de Cruys, S. (2020). Precise Worlds for
Certain Minds: An Ecological Perspective on the Relational Self in Autism. Topoi.
https://doi.org/10.1007/s11245-018-9546-4.
Curry, G. and Ravenscroft, I. (2002). Recreative Minds. Oxford University Press,
Oxford. doi:10.1093/acprof:oso/9780198238089.001.0001

Penultimate draft. Accepted for publication in Topoi.
Please cite the published version.
24
De Jaegher, H. (2013). Embodiment and sense-making in autism. Front. Integr.
Neurosci. 7:15. doi: 10.3389/fnint.2013.00015
Di Paolo, E., & Thompson, E. (2014). The enactive approach. In L. Shapiro (Ed.),
Routledge handbooks in philosophy. The Routledge handbook of embodied cognition (p. 68–
78). Routledge/Taylor & Francis Group
Di Paolo, E., and De Jaegher, H. (2012). The interactive brain hypothesis. Front.
Hum. Neurosci. 6:163. doi: 10.3389/fnhum.2012.00163
Di Paolo, E., Buhrmann, T., Barandiaran, X., (2017) Sensorimotor Life: An Enactive
Proposal. Oxford University Press
Di Paolo, E., Cuffari, E. C., & De Jaegher, H. (2018). Linguistic Bodies: The
Continuity Between Life and Language. Cambridge, MA, USA: MIT Press.
Frith, U. (1996). Cognitive explanations of autism. Acta Paediatr. 416, 63–68. doi:
10.1111/j.1651-2227.1996.tb14280.x
Fuchs, T., and De Jaegher, H. (2009). Enactive intersubjectivity: participatory sense-
making and mutual incorporation. Phenomenol. Cogn. Sci. 8, 465–486. doi: 10.1007/s11097-
009-9136-4
Gallagher, S. (2004). Understanding Interpersonal Problems in Autism. Philosophy,
Psychiatry, and Psychology 11 (3):199-217.
Gallagher, S. (2017). Enactivist interventions: Rethinking the mind. Oxford University
Press.
Gibson, J. J. (1986). The Ecological Approach to Visual Perception. Psychology Press
, London.
Happé, F., and Frith, U. (2006). The weak coherence account: detail-focused cognitive
style in autism spectrum disorders. J. Autism Dev. Disord. 36, 5–25. doi: 10.1007/s10803-005-
0039-0
Hens, K., and Langenberg, R. (2018). Experiences of Adults Following an Autism
Diagnosis. Cham: Springer. doi: 10.1007/978-3-319-97973-1
Hipólito, I, Hutto, D, & Chown, N. (2020). Understanding autistic individuals.
Cognitive diversity not theoretical deficit. In Rosqvist, H. B., Chown, N., and Stenning, A (eds),
Neurodiversity Studies. A New Critical Paradigm. Routledge, London.
https://doi.org/10.4324/9780429322297
Hohwy, J. (2013). The predictive mind. Oxford University Press.

Penultimate draft. Accepted for publication in Topoi.
Please cite the published version.
25
Hutto, D. D., & Myin, E. (2013). Radicalizing enactivism: Basic minds without
content. Mit Press.
Hutto, D. D., & Myin, E. (2017). Evolving enactivism: Basic minds meet content. MIT
press.
Hutto, Daniel D. (2003). Folk psychological narratives and the case of autism.
_Philosophical Papers_ 32 (3):345-361.
Jaswal, V., & Akhtar, N. (2019). Being versus appearing socially uninterested:
Challenging assumptions about social motivation in autism. Behavioral and Brain Sciences,
42, E82. doi:10.1017/S0140525X18001826
Jones, R. M., Southerland, A., Hamo, A., Carberry, C., Bridges, C., Nay, S., Stubbs, E.,
Komarow, E., Washington, C., Rehg, J. M., Lord, C., & Rozga, A. (2017). Increased Eye
Contact During Conversation Compared to Play in Children With Autism. Journal of Autism
and Developmental Disorders, 47(3), 607. https://doi.org/10.1007/s10803-016-2981-4
Kenny, L., Hattersley, C., Molins, B., Buckley, C., Povey, C., and Pellicano, E. (2016).
Which terms should be used to describe autism? Perspectives from the UK Autism Community.
Autism 20, 442–462. doi: 10.1177/13623613155 88200
Krueger, Joel & Maiese, Michelle (2018). Mental institutions, habits of mind, and an
extended approach to autism. Thaumàzein 6:10-41.
Lyons V., and Fitzgerald M., (2012). Critical Evaluation of the Concept of Autistic
Creativity, In Fitzgerald, M. (ed) Recent Advances in Autism Spectrum Disorders - Volume I,
IntechOpen. doi: 10.5772/54465
Markram, K., and Markram, H. (2010). The intense world theory – a unifying theory of
the neurobiology of autism. Front. Hum. Neurosci. 4:224. doi: 10.3389/ fnhum.2010.00224
Maturana, H. R., & Varela, F. J. (1980). Autopoiesis and Cognition—The Realization
of the Living, ser. Boston Studies on the Philosophy of Science. Dordrecht, the Netherlands.
McGeer, V. (2007). Why neuroscience matters to cognitive neuropsychology. Synthese
159, 347–371. doi: 10.1007/s11229-007-9234-1
McGeer, V. (2019). Scaffolding agency: A proleptic account of the reactive attitudes.
Eur J Philos.; 27: 301– 323. https://doi.org/10.1111/ejop.12408
Milton, D. (2017). A Mismatch of Salience: Explorations of the Nature of Autism from
Theory to Practice. West Sussex: Pavillion.

Penultimate draft. Accepted for publication in Topoi.
Please cite the published version.
26
Mottron, L., Dawson, M., Soulières, I., Hubert, B., and Burack, J. (2006). Enhanced
perceptual functioning in autism: an update, and eight principles of autistic perception. J.
Autism Dev. Disord. 36, 27–43. doi: 10.1007/s10803-005- 0040-7
Murray, D. (2018) Monotropism: An interest-based account of autism, Encyclopedia of
autism spectrum disorders, Springer New York, 978–981
Noë, A. (2004). Action in perception. MIT press.
O’Regan, J. K., & Noe, A. (2001). A sensorimotor account of vision and visual
consciousness. Behavioral & Brain Sciences, 24(5), 939.
Palmer, C. J., Paton, B., Kirkovski, M., Enticott, P. G., & Hohwy, J. (2015). Context
sensitivity in action decreases along the autism spectrum: a predictive processing perspective.
Proceedings of the Royal Society B: Biological Sciences, 282(1802), 20141557
Pellicano, E. (2013). Sensory symptoms in autism: a blooming, buzzing confusion?
Child Dev. Perspect. 7, 143–148. doi: 10.1111/cdep.12031
Richman, KA, Bidshahri, R. Autism, theory of mind, and the reactive attitudes.
Bioethics. 32: 43– 49. https://doi.org/10.1111/bioe.12370
Van de Cruys, S., Kelsey, P. & Hohwy, J., (2019): Explaining hyper-sensitivity and
hypo-responsivity in autism with a common predictive coding-based mechanism, Cognitive
Neuroscience, DOI: 10.1080/17588928.2019.1594746
Servick, K., (2019, October 1). Echolocation in blind people reveals the brain’s
adaptive powers. Science Mag (retrieved 13th of June 2020),
https://www.sciencemag.org/news/2019/10/echolocation-blind-people-reveals-brain-s-
adaptive-powers
Shoemaker, D. (2015). Responsibility from the Margins, Oxford University Press
Sterelny, K. (2010). Minds: Extended or scaffolded? Phenomenology and the Cognitive
Sciences, 9, 465–481.
Strawson, P. F. (2008). Freedom and Resentment and Other Essays. Routledge.
https://doi.org/10.4324/9780203882566.
Thompson, E. (2007). Mind in life. Harvard University Press.
Valiyamattam, G. J., Katti, H., Chaganti, V. K., O’Haire, M. E., & Sachdeva, V. (2020).
Do Animals Engage Greater Social Attention in Autism? An Eye Tracking Analysis. Frontiers
in Psychology, 11, 1.

Penultimate draft. Accepted for publication in Topoi.
Please cite the published version.
27
Van de Cruys, S., Evers, K., Van der Hallen, R., Van Eylen, L., Boets, B., de-Wit, L., et al.
(2014). Precise minds in uncertain worlds: predictive coding in autism. Psychol. Rev. 121, 649–
675. doi: 10.1037/a0037665
van Es, T. (2020a). Living models or life modelled? On the use of models in the free
energy principle. Adaptive Behavior. https://doi.org/10.1177/1059712320918678
van Es, T. (2020b) Minimizing prediction errors in predictive processing: from
inconsistency to non-representationalism. Phenom Cogn Sci 19, 997–1017.
https://doi.org/10.1007/s11097-019-09649-y
Van Goidsenhoven, L. and De Schauwer E. (2020), Listening beyond words: swinging
together, Scandinavian journal of disability research, p. 330-339
van Grunsven, J. (2020). Perceiving 'Other' Minds: Autism, 4E Cognition, and the Idea
of Neurodiversity. Journal of Consciousness Studies 27 (7-8):115-143
Varela, F. J., Thompson, E., and Rosch, E. (1991). The embodied mind: cognitive
science and human experience. Cambridge, MA: MIT Press.
Vygotsky, L. S. (1978). Mind in society: The development of higher psychological
processes. Cambridge, MA: Harvard University Press.
Wallace, R. (2019). The Moral Nexus. PRINCETON; OXFORD: Princeton University
Press. doi:10.2307/j.ctv3znwhn
Zahnoun, F. (2019) On representation hungry cognition (and why we should stop
feeding it). Synthese. https://doi.org/10.1007/s11229-019-02277-8
Zaidel, A., Goin-Kochel, R. P., & Angelaki, D. E. (2015). Self-motion perception in
autism is compromised by visual noise but integrated optimally across multiple senses.
Proceedings of the National Academy of Sciences, 112(20), 6461-6466