To understand why we lose these arguments, we first have to accept a harsh reality about the general public: A massive percentage of the population is legally illiterate, deeply apathetic, arrogant, and entirely illogical.
According to OECD PIAAC data, roughly one third of the adult population is functionally illiterate or lacks the critical cognitive skills required to parse multi-step logic. Compounding this is profound political apathy; as Achen & Bartels (2016) demonstrated, the vast majority of voters make decisions based entirely on pre-programmed social identities rather than rational policy logic.
We don't have to look far for proof. Look at major geopolitical events. In the 2016 Brexit referendum, 51.9% of the UK voted to leave the EU (UK Electoral Commission, 2016 as cited in Wikipedia) despite overwhelming economic warnings, driven largely by emotion and rhetoric.
When we argue with them, we aren't arguing with a free thinker, assuming what cognitive psychologist Kahneman describes as someone engaging in high-effort, analytical 'System 2' cognition, or what Jonathan Baron (1993) defines as possessing the 'active open-mindedness' required to update beliefs based on new evidence.
Instead, we are probably arguing with what philosopher David Chalmers (1996) would classify as a "p-zombie": a being that simulates logical debate without the conscious, critical reflection behind it. See wikipedia
A Conversation on "Is Racism Real at Uni?"
This is a real conversation I had regarding the 15% grade gap between Asian and white students on subjective continuous assessments. I'd like to revisit this and reflect on how I fell into his traps.
The Demand for Data?
Me: I really don't want to take this course with continuous assessments. I can tell in most cases, Asian students are being constantly disadvantaged in continuous assessments, scoring like 15% lower than white students.
Tech Bro: Where's that data from?
Me: From what I've seen. Like, across the classes I've taken.
Tech Bro: Do you have the exact data? Like, actual numbers, a study, anything?
Deflection here: Notice the immediate pivot. Asking for the mathematical proof of a microaggression or asking a student for a university-wide audit on a specific rubric is like asking for the exact gravitational math of a paper cut. It's designed to be impossible to provide on the spot.
The "Burden of Proof", again
Me: It's just a pattern I've noticed. But do you see a pattern here? It seems like something...
Tech Bro: Look, I get that you're frustrated, but calling it systemic bias without a massive dataset is just playing the victim card. If you don't have the numbers, you can't prove it's racism. Burden of proof is on you.
Me: How is that not racist though? I mean, if a specific group is consistently getting marked down-
Tech Bro: How is it racist? You're the one making the claim. Causation isn't correlation.
Me: You meant "correlation isn't causation", right? Do you know what implicit bias actually means? One of the characteristics of racism is that you cannot always prove it mathematically, but rather you look at the systemic outcomes. Like when the system is marking down a group and nobody has the data for it because nobody's collecting it. So if somebody just walked up and said "nihao" to me unprompted, is that racist or just-
Tech Bro: That's just what you saw. The sample size is too small. Causation is not correlation. You can't just assume it's systemic bias based on a handful of grades. Data according to who? Burden of proof relies on you again.
The J.A.D.E. Mistake: Here is where I lost. I started to Justify, Argue, Defend, and Explain (J.A.D.E.). I accepted his premise that I needed numbers. I let him use thought-terminating clichés (Lifton, 1961, as cited in Wikipedia) like "correlation isn't causation" on a loop. I was trying to explain sociology; he was just trying to exhaust me.
Deflection
Me: The Asian student population at UniSA STEM only takes up like 15% of all students. Yes, the sample sise isn't huge, but I've witnessed it with every single Asian student I know. What happened to me was that other Asian students and I consistently underperformed by 15% on continuous assessments, but outperformed the class average by 20% in exam-heavy courses.
Tech Bro: Again, that's just what you saw. A huge chunk of the Asian students are international, and honestly, they just have a massive language barrier. If they are bad at communication, obviously they're going to get lower grades on continuous assessments.
Me: I don't disagree that international students do struggle with English. But if you look at overall academic performance and see Aboriginal Australians doing worse than us, then just blame it on their background - how is that not racist?
Tech Bro: But that is a reality though. If you come from a worse high school or have a massive language barrier, obviously you're going to struggle more. It's an education and language issue, not a race issue. Especially in software engineering, continuous assessment is grading teamwork and communication. If you can't communicate your ideas clearly in a group project or presentation, you deserve a lower grade.
The Spec vs. Assimilation argument: He uses a real, separate issue (ESL language barriers) to completely overwrite the existence of implicit bias. He claims the rubric is "objective." But if a student's code documentation is accurate, and they are marked down because they don't sound like a local, you aren't grading engineering. You're grading assimilation.
Deflection Pt.2
Me: No, I get that. Communication is important. I'm not saying it shouldn't be graded at all. I just think are you sure the way it's being graded is totally fair? Like, maybe the rubric is being applied differently depending on who's being assessed?
Tech Bro: Look, it's not a biased metric, it's just reality. If they can't speak English well enough to present their software architecture clearly, they get a lower grade. That's not the training data being biased; that's just them failing the spec. It's objective.
Me: No, even if the spec is "technically objective" on paper, the human being applying it isn't a machine. A grader can still use "communication skills" as a reason to mark someone down when really they just look different, which literally happened to me. I had to ask for a remark to bypass the bias, and my grade immediately went up.
Tech Bro: Right, but if there's no evidence, and the rubric is objective, and the language barrier is a real thing, then what exactly are you even arguing about? You're basically saying "trust me, I feel it" and expecting me to accept that as proof of systemic racism.
Me: Are you sure racism and bias do nothing here?
Tech Bro: And I'm saying without data, it's just a feeling.
Me: So what, I'm supposed to just ignore what I've been watching for years?
Tech Bro: I'm just saying there's plenty of non-racist reasons for a grade gap.
Me: Whatever, okay, call me a conspiracy theorist but it is actually happening and I can't unsee from what I saw.
Tech Bro: ...Okay, whatever. Anyway, I gotta get back to this code.
Racism Exists Everywhere, With or Without Data
It is quite sad that we need a spreadsheet for someone to believe a minority group's experience, and that is exactly how systemic bias survives. But if they demand the macroscopic data and the theory, we have both.
The Intersubjective Illusion
When someone claims a subjective rubric like "communication skills" is objective, they are committing a fundamental epistemological error: they are conflating objectivity with intersubjectivity.
Objective reality exists independently of human observation (e.g., the code compiles, or it throws an error). However, "professionalism" is intersubjective—it is a shared consensus constructed by a specific dominant culture. Because the dominant demographic in STEM shares the exact same intersubjective norms, they mistake their localised echo chamber for an objective law of nature.
The Demographic Myth of the "Default"
They also operate on the heuristic that their own demographic is the baseline of humanity. Statistically, this is false.
The 2011 National Geographic global average person was a 28-year-old Han Chinese man. The 2026 ABS census average Australian is a 40-year-old woman in healthcare. The young, white, male software engineer is a demographic outlier. When minority students are marked down, they aren't failing an "objective spec"—they are being penalised for deviating from the grader's specific cultural dialect.
The Academic Proof of Implicit Bias
This intersubjective bias is mathematically documented across all professional metrics. Look at the foundational CV experiments by Bertrand & Mullainathan (2004), replicated by Wood et al. (2009, as cited in Tate & Damien, 2018), which proved that identical resumes with minority names receive significantly fewer callbacks.
Look at the Bamboo Ceiling phenomenon, where Asian professionals are statistically overrepresented in entry-level technical roles but invisibly barred from leadership. Research on grading bias (Malouff & Thorsteinsson, 2016) strongly suggests that when rubrics rely on subjective criteria—such as evaluating 'communication'—graders are much more susceptible to projecting implicit biases onto minority students
The Institutional Reality
Finally, look at the lived reality on our own campuses. According to the 2026 Human Rights Commission’s Respect@Uni report detailing systemic institutional blind spots, over 80% of Chinese, Northeast Asian, and First Nations respondents report experiencing racism at university.
Yet, this reality remains invisible to the dominant group because the system actively suppresses it: only 6% of victims ever file complaints due to fear of consequences, and up to 80% report profound dissatisfaction with university complaint processes. It is not a "feeling" or a "small sample sise." It is a statistically proven, institutional failure perfectly disguised as objective reality.
A Theoretical (and Theatrical) Analysis
As psychology students, we know that arguing with somebody triggers our fight-or-flight response. As Naomi Eisenberger's (2003) fMRI studies proved, our dACC (dorsal anterior cingulate cortex) processes social conflict and rejection exactly the same way it processes physical pain.
When this happens, the brain shuts down its declarative reasoning centers. If we try to logic our way out, we will just lose, stop thinking clearly, stumble over our own words, and lose.
When I analyse the transcript, my mistakes become obvious: I fell into the J.A.D.E. Trap, a behavioral anti-pattern identified by Al-Anon (as cited in PsychCentral):
- Justify our observation.
- Argue their logic.
- Defend our reality.
- Explain our feelings.
The second I started defending my official university remark against his claim that it was "just a feeling," I gave him the power of the judge.
Which leads to the tactics they use to make us J.A.D.E., while I didn't even notice they are just running on a loop:
- "That only happens to you."
- "Show me the data."
- "That isn't racism, it's just [X]."
- "The burden of proof is on you."
How to Do Better: Translating Declarative to Procedural Knowledge
1. Frame Their "Fact" as an "Option"
1.1 The Visual Metaphor: The Restaurant
When we are stressed, we won't remember the phrase "frame their idea as an option." Instead, use a single visual metaphor that is impossible to forget: The Restaurant.
When the Tech Bro speaks, imagine he is a waiter handing us a menu.
- A waiter cannot force us to eat the fish.
- They are just presenting an option.
- Our only job is to look at the menu, say, "No thanks," and order what we actually want.
During our 3-second pause, ask ourselves: "What's cheaper?"
People usually push an "option" on us because it is the easiest route for them.
According to Fiske & Taylor’s (1991) 'Cognitive Miser' theory, the human brain relies on heuristics because critical thinking burns too many calories.
It is metabolically and psychologically "cheaper" for the Tech Bro's ego to run a script blaming international students' English than to expend the high cognitive effort required to dismantle his own systemic privilege.
Once we realise they are just handing us a "menu" to make their own lives easier, we repeat their option out loud and reject it.
- He says: "It's just objective grading based on communication skills. That's why the grades are lower."
- Our Brain: Pause. He's handing me a menu. His purpose is to protect his worldview.
- We (Framing it as an option): "Look, if you want to operate on the theory that an entire demographic just happens to be worse at the coursework, that is one way to look at it." (We just turned his 'reality' into a 'theory'). "But I'm not buying it. It's intellectually lazy."
By calling their demands "solutions," "theories," or "options," we refuse to accept their premise as reality.
1.2 The Workplace Bully-Calming Method
Once we have successfully reframed their "objective fact" into a mere theory or menu option, we can use the workplace bully-calming method to completely shut down the debate.
Because we cannot out-logic someone who is protecting their ego, our best weapon is profound dismissal. Instead of arguing the mechanics of their theory, we demote their grand facts into mere opinions and politely permit them to hold them.
Our response: "It is okay for you to think that way."
Now, we are simply treating their deeply held "objective reality" as just another harmless, subjective opinion, which means they have nothing to push back against.
2. Force the Alternative Hypothesis (Throw the Ball Back)
If we do decide to engage, never defend our data. Make them defend the void they just created. If they claim it's an "objective rubric" and not bias, force them to say exactly what they think is causing the gap.
Our response: "Are they actually bad at technical communication, or do they just have an accent you don't like? Because if their code documentation is perfectly accurate, but we mark them down because they don't sound like a local, we aren't grading engineering. We're grading assimilation. That is the definition of systemic bias."
This throws the ball entirely back into their court. We are no longer proving that racism exists; they now have to prove that they aren't grading based on assimilation.
3. Defeating Dismissive Gaslighting Traps
Arrogant debaters will often try to dismiss our reality without actually having to debate the facts. They do this by attempting to demote our argument or positioning themselves as the "adult in the room." We have to refuse the demotion.
3.1 The Objectivity Gaslight
This is a classic dismissal where they tell us: "That's just a feeling."
This is an "Appeal to Stone" (Argumentum ad lapidem): a logical fallacy where a person dismisses a claim as absurd without providing any proof of its absurdity. By reframing our observed data as a mere "feeling," they demote us from a logical observer to an emotional victim.
Our response: "Calling it a 'feeling' is just a convenient way for you to dismiss it. Just because you haven't observed it doesn't make my observation an emotion"
3.2 The "Charlie Kirk" Fake Magnanimity Trap
Often, sensational debaters will use a faux-polite dismissal made famous by commentators like Charlie Kirk: "I don't know where you got that data from, but that's okay."
This is a direct cousin of "the dataset is too small" or "correlation isn't causation." It is designed to make them look reasonable and magnanimous while entirely discrediting our premise without having to provide a counter-argument. They are graciously indulging a "naive" claim.
How to defeat it: Instead of accepting their fake grace, mirror their fake magnanimity right back at them or scramble to cite our sources (which instantly triggers J.A.D.E.).
The trick is to start our response with "It's okay that you...":
- "It's okay that you aren't familiar with the research, but your lack of knowledge doesn't make the data less real."
4. The Meta-Callout (If We Fall Into Their Trap)
But what happens if we mess up? What if our amygdala hijacks us, we take the bait, and we suddenly realise we've spent five minutes arguing about syllabus design instead of racism?
We are stuck in their loop. Their algorithm is just playing the hits: "Show me the data", "Burden of proof", etc.
When we realise we are trapped, we use the Meta-Callout. Stop arguing the topic and call out the mechanic they are using to deflect:
- "Do we notice what we're doing right now? Every single time I point out racial bias, we immediately change the subject to syllabus design. Stop hiding behind the rubric and just answer the question: why are we so desperate to pretend bias doesn't exist here?"
5. Breaking the Loop: Grey Rocking & Tone Policing
Eventually, we will realise they aren't actually having a conversation with us, they are just running a script on a loop to exhaust us. When they are stuck in this loop, we do not try to out-logic them. We use the Grey Rock method:
The goal is to become as uninteresting and unresponsive as a grey rock, starving them of the debate they are trying to provoke. Do not defend ourselves. Combine the Grey Rock method with framing their claim as an opinion, give them absolutely nothing to push back against, and shut the door:
- "If you say so."
- "You're entitled to your opinion."
- "Because I have zero interest in debating my reality with you. Are we done here?"
5.1 Surviving Tone Policing
If we successfully block their script, they will almost always resort to one final tactic: tone policing
They will say: "Why are you so upset?" or "You're getting too defensive" or "You're taking this too personally."
Historically, tone policing is a derailment tactic used by dominant groups to preserve privilege (Bailey, 2018; APA Blog Woman in Philosophy: "Tone-Policing and the Assertion of Authority"), dismiss marginalised realities, and regain control of the room.
Psychologically, it often triggers DARVO (Deny, Attack, Reverse Victim and Offender). By accusing us of being "hostile" or "emotional," they attempt to paint themselves as the rational victim of our unreasonable anger. They rely on the false dichotomy that emotion and logic cannot coexist. By labeling our response as "emotional," they automatically restore themselves as the objective, logical authority.
Therefore, saying "I'm not upset!" will lose control of the conversation, trigger the J.A.D.E. response, and force you to perform emotional labor to soothe their ego.
Instead, we lean into the Grey Rock method and grant them permission to misinterpret our tone and expose why they need to view us as angry.
- I'm completely calm. But if my directness makes you uncomfortable, you can end the conversation here."
- It is okay for you to think I'm taking this personally. Pointing out systemic bias usually makes people uncomfortable, and we can just end the conversation here.
And then, we simply walk away.