It Rarely Depends, If Ever

I saw a reel on Instagram where an interviewer asks a med student: "How is social life in med school?" The student pauses, then delivers what they clearly believe is a thoughtful response: "It depends. It differs from person to person." They spend the next few minutes explaining the various ways different personalities might experience med school differently. They go on about introverts versus extroverts, organized planners versus procrastinators, people who prioritize grades versus those who don't.

The student thinks they're being nuanced. They're actually dodging the question entirely.

The Question You Were Actually Asked

When someone asks "how is social life in med school," they're not conducting a survey on human personality variations. They already know that introverts exist. They're aware that some people prioritize studying while others prioritize socializing. None of this is news, and none of it is specific to medical school.

What they're really asking is comparative: How does med school compare to other contexts?

The implicit baseline is right there in the question. They want to know if med school imposes unique constraints on social life: More rigid schedules than law school, heavier workloads than engineering, less free time than liberal arts programs.

They want to know what changed when you entered med school. Did your Thursday nights disappear? Did your friend group shrink? Can you still grab spontaneous dinners, or does everything require two weeks' advance planning around anatomy lab schedules?

These are answerable questions about the structural realities of medical education. But instead of answering them, the student retreated into the safest possible territory: acknowledging that people are different.

The "It Depends" Reflex

"It depends" has become the thinking person's default. It signals "sophistication" in that, look how aware I am of nuance and complexity. It protects you from being wrong - after all, you're not making any concrete claims. It sounds humble, as if you're refusing to oversimplify a complicated reality.

But most of the time, it's intellectual cowardice.

The problem isn't that "it depends" is technically incorrect. The problem is that it's always technically correct, which makes it functionally useless. Yes, individual experiences vary. Yes, context matters. Yes, exceptions exist. These truths are so universal they add nothing to any specific conversation. When you lead with "it depends," you're not providing nuance as you might believe. What you're doing is just avoiding analysis.

Real thinking begins where "it depends" ends. It begins when you ask: Okay, but what are the patterns? What's the central tendency? How does this situation typically play out?

What Comparative Analysis Actually Looks Like

Answering the med school question properly requires you to do something the student failed to do: comparative analysis.

You'd need to think about structural constraints. Med school typically demands 60-80 hours of combined class and study time per week, compared to 40-50 for most undergraduate programs. Clinical rotations often include overnight shifts and weekend duties that simply don't exist in law school. The sheer volume of memorization required is qualitatively different from the workload in philosophy or political science.

You'd need to think about cohort effects. Med students are often slightly older, more career-focused, and selected specifically for their ability to handle extreme academic pressure. This creates different social dynamics than you'd find in a more diverse undergraduate population.

You'd need to think about institutional culture. Many medical schools have peer support networks precisely because the workload is crushing - a different social pattern than the more individualistic culture of some graduate programs.

None of this erases individual variation. The introverted med student will still have a different experience than the extroverted one. But these structural realities do create patterns that can be meaningfully described and compared. The question IS answerable. You just have to be willing to answer it.

Why We Avoid Comparison

I think "it depends" forms because comparative analysis is genuinely difficult and genuinely uncomfortable for most people.

It's difficult because it requires you to identify relevant comparison points, understand what makes situations truly analogous or different, and synthesize your observations into generalizations that hold up under scrutiny. It's easier to notice that people vary than to identify how contexts vary.

It's uncomfortable because making comparisons means making judgments. Saying "med school has less social life than law school" feels like you're diminishing med school, or attacking people who chose it, or claiming some kind of superiority. We've become so allergic to value judgments that we'd rather refuse to compare at all than risk offending someone.

But here's the issue: comparison isn't attack. Description isn't prescription. Saying that med school typically leaves less time for socializing than an English literature program is not saying that med students are sad or that literature students are frivolous. It's just... describing reality. And if we can't describe reality because we're terrified of the emotional valence someone might assign to that description, we've surrendered our ability to think clearly about anything.

The Illusion of Infinite Variation

There's something almost superstitious about how we treat human variation. We've convinced ourselves that people are so magnificently different, so irreducibly complex, that any attempt to identify patterns is doomed to failure. It's a flattering story. We all like to think we're special, but it's fundamentally bordering on becoming a logical fallacy.

The fallacy works like this: because individual differences exist, and because we can always point to someone who defies any generalization, we conclude that generalization itself is impossible or inappropriate. However this is like saying that because some people are seven feet tall, we can't meaningfully discuss average human height. Or because some people survive falling from airplanes, we can't say that falls from great heights are generally fatal.

The error is treating the existence of variation as evidence that the variation is too large to analyze. But variation itself can be measured, quantified, and bounded. When we actually look at the data, we find that human variation exists within remarkably consistent ranges. We're not as different as we think we are.

Watch what happens in a debate when someone asks: "Does social media make people more anxious?"

The "it depends" respondent will immediately reach for individual variation: "Well, some people use social media mindfully and feel connected, while others doomscroll and feel terrible. Some people have strong real-world support systems that buffer them, others are isolated. Some people are predisposed to anxiety, others aren't."

All technically true. All completely beside the point.

The person asking the question already knows that humans vary. What they want to know is whether social media, as a systematic force, shifts the distribution of anxiety in a population. Does the average user experience more anxiety? Do clinical anxiety diagnoses increase with social media adoption? When people quit social media, do anxiety symptoms tend to improve?

These are empirically answerable questions. Multiple studies have found that yes, social media use correlates with increased anxiety, particularly in adolescents. Yes, experimental studies where people reduce social media use show decreased anxiety symptoms. The effect sizes aren't massive (we're talking about shifts of maybe half a standard deviation) but they're real, measurable, and consistent across populations.

Saying "it depends on the person" doesn't make you sophisticated. On the contrary, it makes you someone who mistakes the existence of outliers for the absence of patterns. The debate moves forward when someone is willing to say: "Yes, on average, with documented effect sizes, social media increases anxiety for most users, while a smaller subset experiences neutral or positive effects." That's an answer. That's something we can actually discuss, challenge, or build upon.

How the Rest of the World Actually Works

Here's what's fascinating: every field that takes itself seriously has already solved this problem. They just don't tell specifically.

The United Nations ranks 193 countries by Human Development Index. They don't say "well, it depends on which citizens you ask." They aggregate data on life expectancy, education, and per capita income, and they produce a number. Norway ranks 1st. South Sudan ranks 193rd. Yes, there are wealthy people in South Sudan and struggling people in Norway. No, this doesn't make the index meaningless.

Financial analysts use indexes to describe markets containing thousands of individual stocks with wildly different performances. The S&P 500 can be "up" even though hundreds of individual stocks within it are down. Nobody stands up in a board meeting and says "well, it depends on which stock you're looking at" when asked how the market performed. They give you the index, because the index captures something real about central tendency.

Consider the Gini coefficient, which measures income inequality. It reduces the entire income distribution of a nation (every single person's earnings) into a single number between 0 and 1. Does everyone in a country with a Gini coefficient of 0.45 experience inequality identically? Of course not. Does the coefficient still tell you something meaningful about that society's economic structure? Absolutely. South Africa's Gini coefficient of 0.63 isn't invalidated by the existence of poor Norwegians or wealthy South Africans. It describes a pattern that has real consequences for policy, social stability, and human welfare.

Cognitive psychologists measure reaction times in milliseconds. They know your reaction time on any given trial depends on whether you slept well, whether you're caffeinated, whether you're distracted, whether the stimulus was bright or dim. They measure you anyway. They compare you to population norms. They draw conclusions. And somehow, these conclusions replicate across laboratories and cultures and decades, because it turns out human cognition doesn't actually vary as much as we pretend it does.

The Stroop effect (the delay in reaction time when color words are printed in conflicting colors, you might've seen it on social media) occurs in essentially every human tested, across cultures, ages, and contexts. Yes, the magnitude varies. Some people show a 50ms delay, others show 200ms. But nobody shows zero. Nobody shows a facilitation effect instead. The variation exists within bounds narrow enough that we can say: humans have automatic reading processes that interfere with color naming. This is true. It doesn't depend. And no, there's no "Rain Man" that just is REALLY GOOD at this.

Neuroscientists will tell you where Broca's area is. Not where it might be depending on the person. Where it is. Yes, there's some individual variation in exact location. No, this doesn't prevent them from doing brain surgery. The variation exists within such a narrow range that they can make reliable predictions about what will happen if that area is damaged. Damage Broca's area, you get expressive aphasia (difficulty producing speech while comprehension still works). This happens in the vast majority of cases. Not all cases. The vast majority. Which is enough to base an entire subfield of neurology on.

The pattern is everywhere once you notice it: Fields that actually have to produce actionable knowledge gave up on "it depends" decades ago. They found ways to measure central tendency, to quantify variation, to make meaningful comparisons despite individual differences. They had to, because "it depends" doesn't cure diseases or predict market crashes or design bridges that stay up.

The Debate Stage Defense

You see the "it depends" strategy deployed most cynically in debates and political discourse, where it serves as an all-purpose deflection.

Ask a politician: "Will this policy reduce unemployment?"

Watch them respond: "Well, it depends on so many factors! The global economy, technological change, individual work ethic, regional industries, demographic trends..."

They're not wrong. Those factors exist. But notice what they've done: They've made the question unanswerable by expanding the scope of "depends" to include literally everything that could possibly affect the outcome. It's outright strategic ambiguity.

The honest answer would be: "Based on similar policies implemented in comparable economies, we'd expect a reduction of approximately 1-2 percentage points in unemployment over two years, with larger effects in manufacturing regions and smaller effects in service economies. Of course, external shocks could alter this, but that's our baseline projection given historical data."

One answer tells you nothing. The other gives you an expected range, acknowledges uncertainty, and still makes a falsifiable claim. One is hiding behind variation. The other is doing analysis.

Or watch what happens when someone asks: "Is remote work more productive than office work?"

The "it depends" debater will catalog individual differences: "Some people are self-motivated and thrive at home, others need structure and struggle with distractions. Some jobs require collaboration, others are heads-down work. Some companies have good remote infrastructure, others don't."

Fine. Now what? You've listed variables. You haven't synthesized them into an answer. You haven't told anyone what the actual research shows: that for most knowledge workers, productivity either stays constant or slightly increases with remote work.

The "it depends" response feels smart because it demonstrates awareness of complexity. But awareness of complexity isn't the same as managing complexity. Real analytical thinking takes all those variables and synthesizes them into probabilistic claims: "For roles X and Y, under conditions A and B, we typically see outcome C, with variance range D." That's what adults do when they need to make actual decisions.

The Spectrum Excuse

But human behavior exists on a spectrum, you might object. People really are different.

True. And irrelevant.

Yes, humans exist on behavioral spectrums. But these spectrums have limits - biological, physical limits - that constrain the range of possible variation. You're not going to find a human who constantly sleeps two hours a night and feels great, no matter how much they insist they're just built different. You're not going to find someone whose reaction time is negative. You're not going to find a population that thrives on 500 calories a day.

The human body temperature spectrum runs from about 97°F to 99°F for healthy individuals. Yes, there's variation. No, this doesn't mean we can't identify fever. When someone's temperature hits 103°F, we don't say "well, body temperature depends on the person." We say "you have a fever" and we treat it.

More importantly, most people aren't at the edges of these spectrums. They're clustered in the middle. This is what normal distribution means, it's not a value judgment, it's a statistical reality. Most people sleep somewhere between six and nine hours. Most people have similar caloric needs within a fairly narrow band. Most people's working memory can hold roughly seven items, plus or minus two.

This clustering isn't coincidental. It reflects the fact that we're the same species, with the same basic neurological architecture, the same evolutionary history, the same physiological constraints. The variation that exists is variation within a tightly bounded system, not variation across an infinite possibility space.

When you say "it depends on the person," you're usually pointing at variation that exists within a range so narrow it doesn't actually change the answer to the question being asked. Yes, some med students manage to maintain active social lives. They're the outliers. The overwhelming majority experience a significant reduction in social activity compared to their previous lives and compared to peers in less demanding programs. This is the pattern, and this is the answer.

Take intelligence testing. Notice what IQ testing reveals about human variation: most people score between 85 and 115. Not between 0 and 200. Between 85 and 115. Sixty-eight percent of all humans cluster in a 30-point band. That's remarkably little variation for something we think of as highly individual.

Or consider height. Adult human height varies from about 148cm to 199cm (4'10" to 6'6") for the vast majority of people: A range of about 50cm, around an average of 170cm (20 inches around an average of 5'7"). That's a coefficient of variation of about 5%. We're incredibly similar in height, and yet we still talk about tall people and short people as though they're different categories of being. Imagine how little actual variation there is in behaviors we think are wildly different.

Acknowledging that outliers exist doesn't make the pattern disappear. It just makes you look like you're afraid to name it. Or worse yet, someone who can't see the forest for the trees.

When Dependence Doesn't Matter

Even when things genuinely depend on context, they often depend within such a constrained range that the dependence itself becomes trivial.

Take pain tolerance. Yes, pain tolerance varies between individuals. Some people cry at paper cuts, others walk on broken bones. Genuine variation exists. But if I hit you with a hammer, you're going to feel pain. I don't need to qualify this with "well, it depends on the person." The dependence exists within a range that doesn't change the fundamental answer: Hammer hurts human.

The difference between the person with the highest pain tolerance and the person with the lowest pain tolerance is maybe a factor of two or three in terms of stimulus intensity required to produce equivalent pain ratings. Not a factor of a hundred. Not a factor of ten. Maybe three. That's the actual range of human variation we're talking about.

Or consider: "Is surgery stressful for patients?"

You could say "it depends." Some people are anxious by nature, others are calm. Some have had good experiences with medical care, others have trauma. Some have strong support systems, others are alone. All true. But surgery remains stressful for the vast, vast majority of people, even those at the "low stress" end of the spectrum. The variation isn't large enough to make "it depends" a better answer than "yes."

Measure cortisol levels before surgery. Measure heart rate variability. Measure self-reported anxiety. You'll find that essentially everyone shows physiological and psychological stress responses. The magnitude can vary, maybe the calmest patient has a 20% cortisol increase while the most anxious has a 200% increase. But they're both stressed. The question "is surgery stressful" has an answer, and that answer is yes.

This is the test: when the range of variation is small compared to the magnitude of the phenomenon you're describing, the dependence doesn't actually matter. If med school reduces available social time from 30 hours a week to 5 hours a week for most students, and the variation is between 3 and 8 hours depending on personality and priorities, you don't lead with "it depends." You lead with "massive reduction," and you note the variation as a footnote.

Here's another one: "Is learning a language as an adult harder than learning one as a child?"

The "it depends" crowd will tell you about motivated adults who become fluent, about children who grow up bilingual effortlessly, about individual learning styles and immersion contexts. But if you actually look at the neuroscience, it's unequivocal: the critical period for language acquisition closes in early adolescence. Adults can absolutely learn languages, but the neuroplasticity advantage children have is not a minor effect. It's not close. Children acquiring language activate different brain regions, achieve native-like pronunciation at vastly higher rates, and internalize grammar implicitly rather than explicitly.

Yes, some adults achieve high proficiency. These are the outliers on the right tail of the distribution, often people with exceptional aptitude or extraordinary dedication. Pointing to them doesn't invalidate the general pattern. The general pattern is: language acquisition is significantly easier for children. This is true across cultures, across languages, across individuals. It doesn't depend as much as you want it to.

The Myth of Unique Preferences

People like to apply "it depends" to basic human preferences and behaviors.

Ask someone: "Do people prefer being treated with respect over being insulted?"

Watch someone actually say: "Well, it depends on the person. Some people actually like insults, they communicate with insults. Some cultures are more direct. Some people appreciate brutal honesty."

Stop. Just stop.

Humans universally prefer being treated well over being treated badly. This is observable reality across every society ever studied. Yes, the expression of respect varies by culture. Yes, some people have higher or lower thresholds for what they consider disrespectful. But nobody actually prefers contempt to regard, cruelty to kindness, humiliation to dignity.

The anthropological record is unambiguous on this. Every human society has concepts for hospitality, fairness, reciprocity, and care. Every language has words for gratitude and appreciation. No culture celebrates betrayal as a virtue or teaches children that being harmed is preferable to being helped. These are concrete human universals that emerge from our social nature as a species.

Or try this one: "Do people enjoy eating delicious food?"

The "it depends" person will say: "Well, some people have eating disorders. Some people are on restrictive diets. Some people have different taste preferences."

All true. All missing the forest for the trees.

Eating disorders are disorders, deviations from normal functioning that cause distress. They don't represent alternative preferences; they represent systems breaking down. Restrictive diets are typically adopted despite preferences, not because of them. You go vegan for ethical reasons or do keto for health reasons, not because vegetables taste better than pizza. And yes, taste preferences vary, some people like spicy food, others don't, but this variation occurs within bounds. Nobody's favorite food is spoiled meat. Nobody's ideal meal is a bowl of sand. Nobody genuinely prefers starvation to satiation.

Humans are attracted to: high-calorie foods, sweet tastes, savory umami flavors, fresh rather than rotten smells. These preferences are hardwired because they kept our ancestors alive. The variation in which specific foods we prefer is real but superficial: You might prefer Thai food while your friend prefers Italian, but you both prefer food that tastes good to food that tastes bad. This isn't profound individuality. It's trivial variation within a universal template.

The same pattern holds for virtually every human preference:

Do people prefer comfort to pain? Yes.
Do people prefer safety to danger? Yes.
Do people prefer having autonomy over being controlled? Yes.
Do people prefer being understood over being misunderstood? Yes.
Do people prefer success over failure in their endeavors? Yes.

These aren't controversial claims. They're so obvious that stating them feels stupid. But next time, watch what happens when you ask these questions in casual conversation, you'll see that suddenly everyone becomes a philosopher of radical individual difference, pointing out edge cases and contextual factors as though they invalidate the entire pattern.

Someone will mention masochists when you say people avoid pain. Someone will cite soldiers when you say people prefer safety. Someone will bring up people who self-sabotage when you say people prefer success. And all of these examples are real. But they're a tiny fraction of human experience, and even within these examples, the underlying preferences often still hold: Masochists typically seek controlled pain in consensual contexts that they can stop, soldiers accept danger for higher values like loyalty or protection, people who self-sabotage usually do so due to conflicting fears rather than actual preference for failure.

The basic structure of human motivation is not mysterious. We approach rewards and avoid punishments. We seek pleasure and avoid pain. We want to be competent, connected, and autonomous. These drives exist in every human in every culture. The specific objects of these drives might vary in a limited spectrum, what counts as rewarding, what relationships look like, what competence means. But the drives themselves don't vary. They can't, because they're built into our neurobiology.

When you say "it depends" about basic human preferences, you're not being thoughtful. You're denying the existence of human nature itself.

The Mathematics We Already Have

Here's something that should embarrass us: we've had the mathematical tools to handle variation for over three hundred years. We just pretend we don't when it's convenient.

The mean and median (concepts so fundamental they're taught to children) exist specifically to extract signal from noise, to find the central tendency when data points vary. When you calculate the average temperature in July, you're not pretending every July day is identical. You're acknowledging variation while still making a meaningful summary statement.

Standard deviation quantifies how much things vary. A standard deviation of 2 tells you something different than a standard deviation of 20. We can measure the spread, describe it precisely, and incorporate it into our conclusions. Saying "the average med student studies 65 hours per week with a standard deviation of 10 hours" is infinitely more useful than saying "it depends on the person."

Statistics as a field exists because variation exists. The entire discipline is a systematic response to the problem of individual differences. And what statistics tells us, consistently, is that despite variation, patterns remain identifiable and useful.

Consider confidence intervals. When a study reports that remote work increases productivity by 5% with a 95% confidence interval of [2%, 8%], they're not saying every single worker becomes exactly 5% more productive. They're giving you a range that captures the uncertainty while still making a concrete, falsifiable claim. This is how adults handle variation. They measure it, bound it, and communicate it clearly.

Or look at regression analysis, which exists specifically to predict outcomes while accounting for multiple sources of variation. You can model how study hours, sleep, prior knowledge, and test anxiety all simultaneously affect exam performance. The model doesn't say everyone with X study hours gets Y grade. It says that on average, controlling for other factors, X study hours predicts Y grade with this much residual variation. That's an answer. That's useful. That's something you can make decisions with.

Now consider Markov chains and decision trees (mathematical frameworks for mapping out complex decisions with multiple possible paths and outcomes). A Markov process can have millions of possible state transitions, and yet we can still calculate expected values, steady-state probabilities, and optimal decision strategies.

Think about what this means. You can have a system where every single decision point branches into multiple possibilities, where the number of total possible paths grows exponentially, where individual trajectories vary wildly..... and you can STILL make meaningful predictions about where the system will end up. We do this for weather prediction, financial modeling, machine learning, natural language processing. The stock market can move in literally infinite directions, and we still calculate expected returns. Speech recognition has to handle infinite possible sentences, and yet Siri usually understands you.

The math doesn't break down because variation exists. The math exists because variation exists. We built these tools specifically to extract meaningful patterns from messy, variable data.

So when someone asks "how long does it take to learn Python" and you say "it depends," you're ignoring centuries of mathematical development that exists specifically to answer questions like this. The real answer is something like: "For someone with no programming experience, achieving basic proficiency typically takes 3-6 months of consistent practice, with about 200-300 hours of active coding."

That's using the mathematical tools we have. That's treating variation as something to measure and communicate, not something to hide behind.

We calculate batting averages even though every at-bat is different. We report average lifespans even though individuals die at different ages. We give expected delivery dates even though traffic varies. We post restaurant wait times even though party sizes differ. We publish graduation rates even though students have different abilities.

In every domain where stakes are real and decisions matter, we've figured out how to give meaningful answers despite variation. It's only in casual conversation, where we're trying to sound thoughtful without doing actual thinking, that we suddenly forget math exists.

Why We Need to Believe It Depends

I think part of the reason we cling to "it depends" is that it's democratizing. If everything depends on individual choices and circumstances, then everyone's experience is equally valid, equally important. No one's reality can be dismissed as atypical or less relevant.

There's something generous in this impulse. But there's also something intellectually dishonest.

Not all experiences are equally representative. The person who smokes two packs a day and lives to 95 exists. They're real. Their experience is valid. But they don't tell us much about whether smoking causes cancer. The cancer patient who never smoked also exists. Also real, also valid. Also not particularly informative about the general relationship between smoking and cancer.

When we refuse to distinguish between typical and atypical, between central tendency and outlier, we're not being inclusive. We're just being incoherent. We're pretending that statistical distributions don't exist, that population-level patterns aren't real, that we can't make meaningful generalizations about anything.

This renders almost all practical knowledge impossible. How do you make medical treatment decisions if every patient is so unique that clinical trials are meaningless? How do you design educational curricula if learning patterns are so individual that pedagogy can't be systematized? How do you build cities if traffic flow is too variable to model?

You can't. And we don't. In practice, every functional institution runs on the assumption that human variation exists within predictable bounds. We just pretend otherwise in casual conversation because it feels less judgmental.

The Way Forward

The next time you find yourself reaching for "it depends," stop. Ask yourself: Am I genuinely pointing out relevant contextual factors that change the answer, or am I just avoiding the work of analysis?

Because most of the time, it doesn't actually depend as much as you think it does. Most of the time, there are patterns. There are central tendencies. There are meaningful differences between situations that can be articulated and compared. Most of the time, the variation exists within bounds narrow enough that the general answer holds.

The med student could have said: "Med school is brutal for social life, honestly. I see friends maybe once a week now instead of every day. The schedule is rigid, the workload is constant, and everyone's exhausted. But I have a friend group, and we study together, we vent together, we occasionally party together. It's different from undergrad. Smaller, more intense, less spontaneous. Yeah, some people manage better than others, but everyone feels it."

That would have been an answer. Specific, comparative, useful. It acknowledges individual variation without hiding behind it. It describes a pattern without claiming it's universal. It respects the intelligence of the person asking the question.

Real sophistication isn't refusing to generalize. It's knowing when generalizations are warranted, understanding their limitations, and having the courage to state them clearly anyway.

Every functioning system in modern civilization runs on our ability to identify patterns, make comparisons, and draw conclusions despite variation. Economics, medicine, engineering, psychology.... None of these fields would exist if "it depends" were an acceptable place to stop thinking.

It's time we recognized that standard.

And no, before you say it: That doesn't depend either.