iiinova — where your digital self keeps evolving

Travis LaCroix @travislacroix.bsky.social · 06/08/2026

I would say so. It's a good paper.

Screen shot of the conclusion of the paper "Strategic Polysemy in AI Discourse" (LaCroix et al. 2026). A section is highlighted, which reads: "Hence, it is worth highlighting that the more things change, the more they stay the same. In the 1970s Drew McDermott criticised the use of “wishful mnemonics”, like goal and understand, to refer to programs and
data structures as theoretically question-begging."

Travis LaCroix @travislacroix.bsky.social · 12/06/2026

10/ Our point is that GenAI is not only a technology problem — it is a question about institutional purpose. If universities want students to take academic integrity seriously, they need to hold themselves to that standard as well. Art: Kate Beaton, Hark! A Vagrant #213.

A black-and-white hand-drawn cartoon showing Jules Verne and Edgar Alan Poe standing inside a hot air balloon labelled “BROS.” They are surrounded by a simple landscape of mountains and sky. The drawing has a rough sketch style with cross-hatching and handwritten lettering. Original art by Kate Beaton, Hark! A Vagrant #213.

010

Travis LaCroix @travislacroix.bsky.social · 12/06/2026

9/ We conclude that GenAI is a threat to academic integrity, because it destabilises the values that integrity presupposes. Hence, if universities wish to hold students accountable in the age of GenAI, they must first hold themselves accountable.

Screenshot of front matter of article, which reads :

[Header] : Higher Education
https://doi.org/10.1007/s10734-026-01706-1.

[Title] : Purpose before policy: academic integrity, generative AI, and
rhetorical stance"

[Authors] : "Tristan B. Taylor and Travis LaCroix"

[Abstract] : "The release of ChatGPT in 2022 sparked polarised responses in higher education, from rapid curricular adoption to warnings about academic misconduct and threats to academic integrity. This article argues that whether generative AI (GenAI) use constitutes misconduct depends on a prior question: what is the university’s purpose? We contend that (apparent) rising GenAI-related misconduct cases reflect structural incoherence in the neo-liberal university, where technological enthusiasm, corporate influence, and policy enforcement often conflict. Such misalignment produces moral and institutional ambiguity, leaving students accountable for behaviours implicitly shaped by the institution itself. Drawing on an historical and rhetorical account of the Anglo-American university model, we analyse mission statements from leading university networks to reveal gaps between how institutional ideals are communicated versus how they are implemented. We conclude that universities cannot credibly enforce integrity standards in the age of AI without first ensuring coherence between their stated missions, pedagogical practices, and approaches to emerging technologies."

Keywords : Higher education · Generative AI · Academic integrity · Academic (dis)honesty · Academic misconduct · Cheating · Large language models · Rhetoric

100

Travis LaCroix @travislacroix.bsky.social · 12/06/2026

8/ We argue that GenAI threatens the higher education, not because its use constitutes misconduct, but because these systems erode academic integrity.

Screenshot of article text reading : "At a basic level, GenAI undermines the pedagogical contract between teachers and students. Instructors assign tasks that purport to teach and evaluate learning, and students respond to those tasks. However, GenAI allows students to meet required outputs without engaging in intended pedagogical processes for which these tasks are designed, calling into question the purpose of the relationship. Part of the issue is that GenAI, framed as an “efficiency enhancer”, privileges quantity over quality, normalises automation as legitimate academic labour, and blurs the line between scholarship and algorithmic output—producing institutional incoherence."

Screenshot of article text reading "Reliance on bullshit is incompatible with the scholarly virtues that academic integrity presupposes. Academic integrity rests on commitments to accuracy, justification, and transparency. LLM outputs violate these norms by simulating expertise without epistemic grounding. Students cannot meaningfully vouch for AI-generated claims. Nonetheless, if a machine-generated essay can achieve a passing grade on an assignment that purports to measure learning, then the assessment itself may reveal less about learning and more about the metric. The process effectively reduces assignments to exercises in plausibility rather than learning, incentivising expediency over engagement. The result is a systemic misalignment in which the very measures used to evaluate learning degenerate into incoherence—a fact made worse because the institution, in promoting AI use, tacitly signals that meeting the output is more important than mastering the process."

Screenshot of article text reading "Evidence also suggests that GenAI impairs learning itself, with automated tools diminishing foundational cognitive capacities, weakening the very competencies that academic integrity presupposes (Appiah, 2025). Students relying on ChatGPT show reduced neural connectivity and poor recall of their own writing (Kosmyna et al., 2025), alongside declines in metacognitive skills and critical thinking, resulting in lower overall reasoning and analytic capacity (Gerlich, 2025). If one of the functions of the university is to transfer knowledge, then it seems counter to this function to promote or permit the use of a technology that has been shown to limit an individual’s ability to learn."

300

Travis LaCroix @travislacroix.bsky.social · 12/06/2026

7/ We highlight that academic integrity is not just about whether an individual student followed the rules. Instead, it as a collective practice — something maintained by students, teachers, institutions, assessment systems, and shared expectations.

Screenshot of article text that reads "There is no universal definition of academic integrity (Hagège, 2023). Nonetheless, the
International Centre for Academic Integrity (ICAI) identifies six core principles—honesty, trust, fairness, respect, responsibility, and courage—without which academic work “loses value and credibility” (ICAI, 2021, 4). The bookending principles of honesty and courage are not to be understated, since honesty is the “indispensable foundation” of integrity (ICAI, 2021, 5) and “only by exercising courage is it possible to create communities that are responsible, respectful, trustworthy, and honest” (ICAI, 2021, 10). Related standards include respect for intellectual property and professional norms of conduct (Balalle & Pannilage, 2025). Crucially, these principles apply to the entire academic community—students and professionals alike (Sbaffi & Zhao, 2022). Academic integrity, however, is not equivalent to the absence of academic misconduct. It entails positive commitments to epistemic virtues—such as sincerity, diligence, accuracy, or justification—and concerns the collective conditions that make learning and reliable knowledge production possible. Integrity cannot be achieved individually; it is a collective endeavour embedded within institutional and social systems (Balalle & Pannilage, 2025)."

100

Travis LaCroix @travislacroix.bsky.social · 12/06/2026

6/ Thinking about the purpose of the institution changes how we think about AI-related misconduct. The question is not : "did a student use AI?" But : "What kind of assignment did we design?" "What learning process were we trying to create?" "What does this assessment actually measure?"

100

Travis LaCroix @travislacroix.bsky.social · 12/06/2026

5/ We also trace how universities have historically served multiple purposes, which have never been perfectly aligned : creating and sharing knowledge, building intellectual communities, supporting social mobility, and contributing to economic systems.

Screenshot of article text, reading: "From its inception, the university has been re-shaped by changing political, religious, economic, and cultural forces, altering its relationship to knowledge creation, preservation, curation, and transmission. Popular narratives oscillate between romanticised visions of timeless sanctuary and cynical depictions of exclusionary gatekeeping. However, both obscure a complex reality: the university is an evolving and contested institution whose purpose has always been subject to debate—particularly with respect to what counts as knowledge, who is authorised to produce it, how it ought to be transmitted, and to whom. Importantly, these debates are not merely epistemic but also material, reflecting struggles over labour, authority, and the distribution of social and economic capital within and beyond the university."

Screenshot of article text that reads "The emergence of GenAI marks another moment of potential transformation, though history reminds us such moments are recurrent. The university has always existed at the intersection of ideals—truth, autonomy, public good—and material constraints shaped by power and economy. Generative AI does not introduce these tensions so much as intensify and expose them, particularly in relation to pedagogy, authorship, and credentialing. Concerns surrounding who is doing the work, what constitutes intellectual labour, and whether a credential has been legitimately earned are not new. These are longstanding questions about academic participation and integrity that GenAI merely reframes. The question posed by AI is not whether the university faces a “new” crisis, but what role it should serve, and whether it can continue to serve that role under new technological conditions. This lineage illustrates that the university is defined not only by its formal functions, but by how it communicates and justifies those functions. Each historical iteration—from mediaeval guild to Humboldtian ideal to neoliberal enterprise—has relied on distinct modes of discourse to legitimate itself. The university’s history is also a history of its rhetorical self-construction. Communication and community, in this sense, are not ends in themselves, but the conditions through which epistemic ecosystems are formed and maintained. Understanding how its function is imagined, communicated, and contested through language leads to the question of the university’s rhetorical constitution—one crucial site of which lies in the values embedded in “academic integrity”."

120

Travis LaCroix @travislacroix.bsky.social · 12/06/2026

4/ Universities often say they value learning, inquiry, and knowledge — but they also reward outputs, metrics, credentials, and efficiency. At the same time, they present AI as a tool for innovation and education, while also treating AI-assisted student work as a threat to academic integrity.

Screenshot or article text reading "constitute academic misconduct, then students’ use of AI systems must, somehow, undermine the university’s function, whatever that function may be. However, this further implies that institutional promotion of GenAI within the pedagogical ecosystem must be incoherent: such a policy would be self-defeating. Yet university administrations appear to endorse both positions: they encourage faculty to integrate AI tools while simultaneously treating GenAI-produced essays as constituting academic misconduct. This inconsistency produces an institutional contradiction. To resolve this incongruity, one should observe that the question of whether students’ use of GenAI in higher education constitutes academic misconduct depends, fundamentally, on what we take the purpose of higher education to be. In this case, how we characterise academic integrity, academic misconduct, or the relationship between the two becomes secondary. Once this purpose is determined, it would be possible to inquire whether the university can continue to satisfy said function or purpose alongside GenAI use. Coherent answers to these questions do not just matter for student behaviour, but also institutional integrity. Indeed, students accused of using GenAI on graded assessments—and subsequently found to have committed academic misconduct—are often unduly sanctioned because senior administrations have failed to communicate the function and purpose of higher education to their stakeholders. When universities fail to clearly articulate their educational goals, or when they adopt policies pertaining to technology use that contradict these goals, they generate an environment of confusion and inconsistency. In this context, it becomes increasingly difficult to hold students solely accountable for choices made within a system that they neither control nor fully understand."

110

Travis LaCroix @travislacroix.bsky.social · 12/06/2026

3/ To explore how universities advertise their purpose, we analyse university mission statements from the Canadian U15, UK Russell Group, and US Ivy League. Institutional language often emphasises "innovation", "excellence", and "impact" without always explaining how those ideals are enacted.

Screenshot of article text reading "(1) knowledge creation and transfer/dissemination, including commitments to research,
teaching quality, and academic freedom;
(2) community and values, often framed in terms of diversity, inclusion, civic responsibility, and ethical leadership; and,
(3) societal or global engagement, expressed through service, innovation, internationalisation, or responsiveness to global challenges."

120

Travis LaCroix @travislacroix.bsky.social · 12/06/2026

2/ Academic “misconduct” only makes sense relative to the goals of the institution. If the purpose of a university is learning, knowledge creation, and participation in a scholarly community, then AI use has to be evaluated by asking whether it supports or undermines those things.

Screenshot of article text that reads "Consequently, institutions have scrambled to develop and implement policies that attempt to grapple with novel technological tools—a fast-paced and moving target. Some institutions have responded with the open adoption and integration of AI tools into curricula and pedagogy. For example, Ohio State University’s AI Fluency initiative seeks to “redefine” learning by embedding AI tools across curricula, including required first-year undergraduate seminars designed to “develop foundational generative AI skills” (The Ohio State University, 2025). At the other end of the spectrum, some groups have issued strong warnings against the uncritical adoption of AI tools in universities, citing potential risks to academic integrity, student learning, and the knowledge ecosystem (Guest et al., 2025). Underlying these increasingly polarised positions is a set of unstated assumptions about the moral and institutional coherence of labelling GenAI-assisted student work as “misconduct”. If the university, as an institution, continues to fulfil its functions in society regardless of students’ use of AI, then such use should not constitute academic misconduct. This holds regardless of what the proper function of the university is—e.g., educational, socio-cultural, political, economic, etc. If the university continues to fulfil its function, then calling GenAI-use “misconduct” is a category mistake. Equivalently: if the employment of these tools does"

100

Travis LaCroix @travislacroix.bsky.social · 12/06/2026

1/ Just published in Higher Ed! We argue that debates about whether GenAI use in universities is "academic misconduct" are asking the wrong question : Before deciding what counts as misconduct, universities need to be clear about what they think universities are for. doi.org/10.1007/s107...

doi.org

100

Travis LaCroix @travislacroix.bsky.social · 28/04/2026

Sad that this paper came out after we wrote ours on strategic polysemy, because "emotion vector" and "functional emotions" are great case studies in *glosslighting*.

Although we have focused on just a few instances of the same phenomenon here, many other terms have similar characteristics as those described above—e.g., consciousness, foundation model, general intelligence, beliefs / preferences / goals, creativity, planning / deliberation, goal-directed behaviour, empathy / affective AI, trustworthiness / honesty, personality, scheming, etc. (See Appendix A.) Across these examples, the same pattern recurs, highlighting how these terms serve a dual purpose. They signal one thing to the public, policymakers, and investors—typically the intuitive, anthropomorphic, or metaphorical meaning—while simultaneously retaining a narrow, technical interpretation to which experts can retreat when challenged. This systematic ambiguity enables, what we call, glosslighting.

Glosslight (Definition):
Glosslighting (verb) is the practice of using technically redefined or polysemous terms to evoke familiar meanings—often emotionally or cognitively powerful ones—while preserving the ability to deny those meanings through retreat into specialised, context-bound reinterpretations.

The rhetorical effects of glosslighting—suggesting familiar meanings while retaining deniability—may arise intentionally, but they can also emerge from (i.e., are a systemic outcome of) the foreseeable interaction between ambiguous terminology (allowing multiple interpretations), and heterogeneous audiences (different groups interpreting the same term differently), combined with institutional incentives (encouraging strategic ambiguity).

020

Travis LaCroix @travislacroix.bsky.social · 28/04/2026

The pattern is reinforced by incentives across research, industry, and media, where intuitive or engaging language travels more easily than precise descriptions. More precise and literal terminology would make it easier to understand what current AI systems actually do, and where their limits are.

110

Travis LaCroix @travislacroix.bsky.social · 28/04/2026

We argue that this kind of ambiguity can make systems seem more capable than they are, make limitations harder to communicate, and affect how responsibility is assigned. (Even when no individual actor is trying to mislead anyone.)

121

Travis LaCroix @travislacroix.bsky.social · 28/04/2026

This ambiguity matters because language shapes public understanding, policy decisions, investment flows, trust in systems, etc. Because these terms circulate beyond research (into media, policy, and industry), their broader meanings often shape how these systems are (mis)understood.

130

Travis LaCroix @travislacroix.bsky.social · 28/04/2026

We introduce the term "glosslighting" to describe this practice in AI discourse: Using familiar words in a technical sense in a way that still evokes their everyday meaning, while keeping the option to fall back on the narrower definition when challenged.

Glosslight (Definition):
Glosslighting (verb) is the practice of using technically redefined or polysemous terms to evoke familiar meanings—often emotionally or cognitively powerful ones—while preserving the ability to deny those meanings through retreat into specialised, context-bound reinterpretations.

This word is a portmanteau of glossary, meaning a list of terms relating to a specific subject, and gaslight, meaning to manipulate someone into questioning their own sanity or powers of reasoning. It is important to note that there is general agreement among philosophers and psychologists that gaslighting need not be intentional or conscious [3, 49, 88] and it is highly unlikely that, where intentional, glosslighting is driven by the same psychological processes as gaslighting behaviour. Indeed, as we shall see below, very different incentives may be at play. It may also be understood as a combination of gloss, meaning to add luster, make shine; the task of advertising, and light, to illuminate, capturing the extent to which discourse is being framed by glossy advertising rather than colder more scientific considerations. Although glosslighting can be understood as the act of glossing over something, this interpretation does not settle whether epistemic manipulation is involved, whereas our definition underscores that the possibility of such manipulation is a core feature of glosslighting.

182

Travis LaCroix @travislacroix.bsky.social · 28/04/2026

AI discourse is full of strategic polysemy (e.g., “Hallucination”, “Reasoning”, “Alignment”, “Agent”). Each of these terms has an entrenched, everyday meaning, but has been appropriated in a narrow technical context, generating ambiguity which, we argue, is not accidental.

181

Travis LaCroix @travislacroix.bsky.social · 28/04/2026

New paper pre-print, accepted at FAccT 2026, wherein @fintanmallory.com , @sashamtl.bsky.social , and I argue that hype-laden AI terms like “hallucination”, “reasoning”, and “agent” are doing strategic, rhetorical work, which we call "glosslighting".

Image of title, authors, and abstract for "Strategic Polysemy in AI Discourse: A Philosophical Analysis of Language, Hype, and Power" (Travis LaCroix, Fintan Mallory, and Sasha Luccioni).

"This paper examines the strategic use of language in contemporary artificial intelligence (AI) discourse, focusing on the wide- spread adoption of metaphorical or colloquial terms like “hallucination”, “chain-of-thought”, “introspection”, “language model”, “alignment”, and “agent”. We argue that many such terms exhibit strategic polysemy: they sustain multiple interpretations simultaneously, combining narrow technical definitions with broader anthropomorphic or common-sense associations. In contemporary AI research and deployment contexts, this semantic flexibility produces significant institutional and discursive effects, shaping how AI systems are understood by researchers, policymakers, funders, and the public. To analyse this phe- nomenon, we introduce the concept of glosslighting: the practice of using technically redefined terms to evoke intuitive—often anthropomorphic or misleading—associations while preserving plausible deniability through restricted technical definitions. Glosslighting enables actors to benefit from the persuasive force of familiar language while maintaining the ability to retreat to narrower definitions when challenged. We argue that this practice contributes to AI hype cycles, facilitates the mobilisation of investment and institutional support, and influences public and policy perceptions of AI systems, while often deflecting epistemic and ethical scrutiny. By examining the linguistic dynamics of glosslighting and strategic polysemy, the paper highlights how language itself functions as a sociotechnical mechanism shaping the development and governance of AI."

76621

Travis LaCroix @travislacroix.bsky.social · 03/03/2026

Me: "I have a new article out!" The AI: "That sucks, bud."

020

Travis LaCroix @travislacroix.bsky.social · 03/03/2026

altmetric sentiment analysis is so funny. It's always like "49% of mentions of your work are STRONG NEGATIVE", and then it's just my own posts on BlueSky.

120

Travis LaCroix @travislacroix.bsky.social · 28/02/2026

Ugh, I know. There was a typo in the .bib file that permeated through the rest of the article. I already contacted the editor to see if we could fix it, so I am hoping it's updated soon!

110

Travis LaCroix @travislacroix.bsky.social · 27/02/2026

Extending my critique, @malinowska.bsky.social et al. argue the persistence of the ToM paradigm in autism research is driven by underdetermination, epistemic-network dynamics, and the institutional payoff of standardization, rather than by its empirical strength. www.tandfonline.com/doi/full/10....

tandfonline.com

Why Does the Theory-of-Mind Paradigm of Autism Persist?

Published in Psychological Inquiry: An International Journal for the Advancement of Psychological Theory (Vol. 36, No. 4, 2025)

180

Travis LaCroix @travislacroix.bsky.social · 27/02/2026

In my response to these commentaries, I shift focus from the "that" question (that ToM-deficit framework is pseudoscience) to the question of why it persists (social dynamics and institutional forces). www.tandfonline.com/doi/full/10....

tandfonline.com

Autism, Theory of Mind, and the Dynamics of Value-Laden Research Programs

This article responds to commentaries on my analysis of the theory-of-mind-deficit explanation of autism, using Lakatos’s methodology of scientific research programs. The commentaries largely agree...

000

Travis LaCroix @travislacroix.bsky.social · 27/02/2026

Baron-Cohen (and friends) disagree. They argue that decades of converging meta-analytic evidence robustly show that autistic people, on average, have dimensional degrees of ToM disability that meaningfully relate to social functioning and should not be dismissed. www.tandfonline.com/doi/full/10....

tandfonline.com

Do Autistic People Have Degrees of Disability in Theory of Mind? The Importance of Meta-Analytic Convergence

Published in Psychological Inquiry: An International Journal for the Advancement of Psychological Theory (Vol. 36, No. 4, 2025)

100

Travis LaCroix @travislacroix.bsky.social · 27/02/2026

Dwyer agrees strong versions are harmful and outdated, but questions labelling weak ToM research as pseudoscientific. It may still offer modest insights—if pursued cautiously, without deficit-based assumptions, and along social change/interdisciplinary dialogue. www.tandfonline.com/doi/full/10....

tandfonline.com

Weak Theories, Research Priority-Setting, and Community Partnership: A Recipe for Success in Polyparadigmatic Autism Science?

Published in Psychological Inquiry: An International Journal for the Advancement of Psychological Theory (Vol. 36, No. 4, 2025)

100

Travis LaCroix @travislacroix.bsky.social · 27/02/2026

Kleberg argues that instead of abandoning ToM research as degenerative, the field should reconceptualise ToM as a heterogeneous set of dissociable processes, investigating how distinct social-cognitive components map onto autism’s diverse developmental profiles www.tandfonline.com/doi/full/10....

tandfonline.com

Staying with the Complexity: Refining Theory of Mind Research in Autism

Published in Psychological Inquiry: An International Journal for the Advancement of Psychological Theory (Vol. 36, No. 4, 2025)

100

Travis LaCroix @travislacroix.bsky.social · 27/02/2026

Rajan supports the claim that the ToM deficit model is degenerating / ad hoc, while further arguing that even reformist alternatives risk presupposing autism as a stable, measurable property and thus leave deeper ontological and political assumptions unexamined. www.tandfonline.com/doi/full/10....

tandfonline.com

Autism Research’s Green Screen: Representationalism and the Political Economy of “Normal” Environments

Published in Psychological Inquiry: An International Journal for the Advancement of Psychological Theory (Vol. 36, No. 4, 2025)

100

Travis LaCroix @travislacroix.bsky.social · 27/02/2026

Bulding on my critique, Gernsbacher argues the persistence of the ToM-deficit account of autism reflects not just empirical weakness but its entanglement with historical narratives, systemic bias, and institutional incentives within psychological science. www.tandfonline.com/doi/full/10....

tandfonline.com

Why the Theory-of-Mind-Deficit Account of Autism Might Persist Despite Evidence to the Contrary

Published in Psychological Inquiry: An International Journal for the Advancement of Psychological Theory (Vol. 36, No. 4, 2025)

110

Travis LaCroix @travislacroix.bsky.social · 27/02/2026

Gough agrees that the ToM-deficit hypothesis should be abandoned, but questions whether pseudoscience is the right framing, arguing instead that ToM research is dehumanising and methodologically deficient regardless of how it is classified in demarcation debates. www.tandfonline.com/doi/full/10....

tandfonline.com

Theory of Mind Research in Autism is Simply Hateful; “Pseudoscience” is Complicated

Published in Psychological Inquiry: An International Journal for the Advancement of Psychological Theory (Vol. 36, No. 4, 2025)

100

Travis LaCroix @travislacroix.bsky.social · 27/02/2026

Further analysis by Morris and Zelazo supports my argument that both strong / weak versions of the ToM deficit hypothesis are empirically unsupported, constituting a degenerating research programme better replaced by neurodiversity-informed alternatives. www.tandfonline.com/doi/full/10....

tandfonline.com

On the Need for Intellectual Humility Regarding Autism: Commentary on Autism and the Pseudoscience of Mind, by Travis LaCroix

Published in Psychological Inquiry: An International Journal for the Advancement of Psychological Theory (Vol. 36, No. 4, 2025)

100

Travis LaCroix @travislacroix.bsky.social · 27/02/2026

The article is accompanied by eight commentaries, and my reply to these commentaries.

110

Travis LaCroix @travislacroix.bsky.social · 27/02/2026

I also argue that progressive alternatives exist (e.g., accounts informed by the neurodiversity paradigm), satisfying Lakatos’s requirement that a research programme be superseded by a more progressive rival. Hence, the theory-of-mind-deficit explanation of autism should be abandoned.

140

Travis LaCroix @travislacroix.bsky.social · 27/02/2026

The implications are nontrivial. The ToM-deficit model has shaped diagnostic discourse, clinical interventions, textbook psychology, legal reasoning, and public understanding of autism A degenerative framework at this level has downstream consequences.

130

Travis LaCroix @travislacroix.bsky.social · 27/02/2026

The paper’s central claim is therefore stronger than prior critiques that “this is bad science.” It is that continued adherence to the theory-of-mind-deficit explanation, in light of its degenerative trajectory, renders the research programme pseudoscientific in Lakatosian terms.

170

Travis LaCroix @travislacroix.bsky.social · 27/02/2026

Rather than abandoning the hypothesis, proponents have: - Expanded the definition of ToM to include its alleged precursors. - Reinterpreted failures as developmental delays. Modified task designs without resolving construct concerns. - Treated autistic self-report as secondary.

160

Travis LaCroix @travislacroix.bsky.social · 27/02/2026

A programme becomes degenerative when: - Auxiliary hypotheses are introduced ad hoc. - Goalposts shift to accommodate counterevidence. - Novel predictions fail to receive empirical support. I argue that the ToM-deficit programme exhibits precisely these features.

182

Travis LaCroix @travislacroix.bsky.social · 27/02/2026

The critique is not limited to empirical shortcomings. Drawing on Imre Lakatos’s Methodology of Scientific Research Programmes, I evaluates the ToM-deficit account at the level of scientific practice. The central question: why has the hypothesis persisted despite sustained anomalies?

171

Travis LaCroix @travislacroix.bsky.social · 27/02/2026

The weak version appears more modest—but collapses under methodological scrutiny. I review evidence that: - ToM tasks (false belief, faux pas, RMET, etc.) often fail to converge. - Replication failures are common. - Linguistic demands confound performance. - Measures lack construct validity.

261

Travis LaCroix @travislacroix.bsky.social · 27/02/2026

This view has already been critiqued repeatedly. See, e.g., Pellicano, 2011 (doi.org/10.1017/CBO9...) and especially Gernsbacher and Yergeau, 2019 (doi.org/10.1037/arc0...)

140

Travis LaCroix @travislacroix.bsky.social · 27/02/2026

Empirically, the strong version fails on standard criteria for explanatory adequacy: - ToM deficits are not universal among autistic people. - They are not unique to autism. - They lack causal precedence. - They do not adequately explain the heterogeneity of autistic traits.

193

Travis LaCroix @travislacroix.bsky.social · 27/02/2026

For decades, the theory-of-mind (ToM) framework has shaped cognitive accounts of autism. Two broad versions persist: - Strong version: autism is fundamentally caused by a ToM deficit. - Weak version: autistic individuals often exhibit ToM difficulties. I argue that both are untenable.

1146

Travis LaCroix @travislacroix.bsky.social · 27/02/2026

Just published in Psychological Inquiry! I offer a sustained philosophical and empirical critique of the theory-of-mind-deficit explanation of autism. The ultimate conclusion is that the research programme has become degenerative—and therefore pseudoscientific. www.tandfonline.com/doi/full/10....

tandfonline.com

Autism and the Pseudoscience of Mind

The theory-of-mind-deficit explanation of autism proposes that autistics lack a theory of mind, that autism comprises a theory-of-mind deficit (strong version); or, that autistics often have diffic...

23421

Travis LaCroix @travislacroix.bsky.social · 20/02/2026

I wrote a whole book that raises / addresses a lot of the questions in this thread re: P-A problems and value alignment! (Including a chapter on who the principal is). broadviewpress.com/product/arti...

broadviewpress.com

Artificial Intelligence and the Value Alignment Problem - Broadview Press

Artificial Intelligence and the Value Alignment Problem -

061

Travis LaCroix @travislacroix.bsky.social · 22/07/2025

I will post more summaries of the results of our search later, but the full article / data summary / analysis can be found (open access!) here: doi.org/10.1007/s112...

First page of article "What do philosophers talk about when they talk about autism", published in Synthese. Information includes the title, doi
(10.1007/s11229-025-05116-1), year (2025), author names (Travis LaCroix, Alexis Amero, Benjamin Sidloski), recieved date (10 October 2024), accepted date (12 June 2025), copyright information (the authors, 2025), abstract ("Several anecdotal claims about the relationship between philosophical discourse and the subject of autism have been forwarded in recent years. This paper seeks to verify or debunk these descriptive claims by carefully examining the philosophical literature on autism. We conduct a comprehensive scoping review to answer the question, what do philosophers talk about when they talk about autism? This empirical work confirms that the philosophy of autism is underdeveloped as a subfield of philosophy. Moreover, the way that philosophers engage with autism is often unreflective and uncritical. As a result, much work in the discipline serves to perpetuate pathologising, dehumanising, and stigmatising misinformation about autistics and autistic behaviour. By highlighting the significant gaps in the philosophical literature on autism, this review aims to deepen our understanding of philosophical thought surrounding autism and contributes to ongoing dialogues pertaining to neurodiversity, madness, and disability rights more generally.") and keywords (Autism · Neurodiversity · The philosophy of autism · Scoping
literature review).

000

Travis LaCroix @travislacroix.bsky.social · 22/07/2025

Examining philosophical works mentioning autism across time, we found: (1) the number of articles has increased significantly in the last decade or two. (2) The rate of change is also trending upward in the last two decades. (3) The majority (> 50%) of the corpus was published in the last decade.

Line graph titled "publications per year." It shows that the number of publications mentioning autism in a given year is trending upward. Each year (1945-2023) is represented on the x-axis, with the number of articles on the y-axis. The maximum is 87 articles in 2021. A dashed line is fitted to the number of articles (blue line), which is steadily increasing in the last 40 years.

Line graph (describing rate of change) and line graph with area underneath filled (describing cumulative distribution), titled "number of publications per year (cumulative distribution)." It shows that the rate of publications mentioning autism in a given year is trending upward. Each year (1945-2023) is represented on the x-axis, with percent (cumulative distribution) on the left y-axis and the rate of change on the right y-axis.

100