Algospeak: Complete Summary of Adam Aleksic’s Analysis of How Social Media is Transforming Language

Introduction: What This Book Is About

Adam Aleksic’s “Algospeak: How Social Media Is Transforming the Future of Language” explores the profound and often unseen ways in which algorithms and social media platforms are reshaping human language. Aleksic, a linguist and content creator known as the “Etymology Nerd,” delves into how online communication drives linguistic innovation, from new slang and grammatical structures to the evolution of accents and the commodification of identity. This book examines how our digital lives influence our everyday speech, often in surprising and paradoxical ways.

The book is for anyone curious about language, social media, and culture, particularly parents, educators, and anyone who uses social media platforms regularly. It offers a fresh perspective on why certain words go viral, how communities form online, and the subtle pressures influencing our speech patterns. Aleksic promises a comprehensive journey into the beautiful, chaotic idiosyncrasies of modern language, revealing how we are constantly adapting and evolving our communication in response to the pervasive influence of algorithms.

By detailing the mechanisms behind viral trends, censorship workarounds, and the creation of online identities, “Algospeak” provides a blueprint for understanding the future of English. It illuminates how memes, metadata, and human psychological tendencies intertwine to create a dynamic linguistic landscape. Readers will gain deep insights into how their own language use is being shaped by the platforms they interact with daily.

1: How to Play Linguistic Whac-A-Mole

This chapter introduces algospeak as a direct response to internet censorship and content moderation, framing it as a linguistic Whac-A-Mole game. Users constantly invent new words or spellings to circumvent algorithmic filters, only for these new terms to eventually be caught and banned, perpetuating a continuous cycle of linguistic creativity and evasion. This phenomenon is not new, but social media platforms accelerate its pace and complexity.

The Evolution of Evasive Language

Evasive language is a long-standing human behavior, with euphemisms for death like “decease” evolving over centuries. The term “unalive” is a modern example of this, gaining prominence as a substitute for “kill” or “commit suicide.” This word’s primary function is euphemistic, making difficult topics more palatable, particularly among middle schoolers. Its casual use, even in formal contexts like student essays, indicates its growing normalization.

Algospeak as a Response to Censorship

The crucial difference for “unalive” is its origin in internet censorship, specifically on TikTok. Starting in 2019, Chinese censorship policies prompted ByteDance to aggressively block content critical of China or sensitive topics like self-harm. This led online creators to adopt alternative words like “unalive” to avoid their videos being removed or suppressed. Viral videos in early 2021 significantly accelerated “unalive’s” usage, eventually leading to its offline adoption.

How Algospeak Differs from Leetspeak

Historically, leetspeak (e.g., “5U1C1D3” for “suicide”) was used to bypass rudimentary text filters. Algospeak, while serving a similar purpose, operates in a more complex environment. Modern AI-powered algorithms are more sophisticated, leading to faster innovation and more guesswork. Creators often experience shadowbanning or suppression without explicit notification, making it harder to know if their content is penalized. This forces creators to err on the side of caution with euphemization.

Bowdlerization and Creative Respellings

The practice of bowdlerization—modifying offensive words—is centuries old. Online, this manifests as creative respellings of swear words (e.g., “btch” for “bitch,” “fck” for “fuck”). This serves not only to avoid censorship but also to add humor. Grawlixes (sequences of graphic characters) in old comics predate these digital forms. Symbolic swearing shifted to replace individual vowels (e.g., “f*ck”) for better intelligibility.

Emojis as Substitutions in Algospeak

Emojis are a prominent form of algospeak for pornographic or sensitive words. The eggplant emoji () for “penis” and peach emoji () for “ass” or “vagina” are common, drawing on visual similarity. Other emojis use acoustic similarity (e.g., corn emoji for “porn,” grape emoji for “rape”). These practices mirror Cockney rhyming slang, showing an age-old process re-emerging in a new digital context.

Minced Oaths and Diminutives

Minced oaths, euphemisms created by slightly altering offensive words (e.g., “heck” for “hell”), are another common algospeak strategy. Words like “seggs” (for “sex”), “nip nops” (for “nipples”), and “peen” (for “penis”) are examples of diminutives that sound smaller, cuter, or less intense, making them more palatable for online use and increasing their viral potential. The phrase “kermit sewerslide” for “commit suicide” is a notable instance.

Initialisms and Metonymy in Evasive Language

Initialisms (e.g., “SA” for sexual assault) have been used for thousands of years to avoid censorship, serving as secret signals within groups. Metonymy, referring to an idea through a closely associated idea (e.g., “White House” for the executive branch), is also increasingly common. Examples include “red zoned” for sexual assault in college, and referring to Hitler as “the top guy of the Germans” to avoid triggering algorithms.

Voldemorting: Avoiding Keywords

Voldemorting is the deliberate avoidance of keywords to prevent content from appearing in searches or being suppressed. This tactic is used to prevent trolls, maintain plausible deniability, and even to signal disrespect (e.g., “cheeto” for Donald Trump). By imposing a taboo on certain words, users reclaim the act of censorship and turn language into an act of resistance.

Algospeak and Geopolitical Conflicts

During the Israel-Gaza war in 2023, social media users developed algospeak like “ğaza” for “Gaza” and “IOF” for “IDF” to discuss the conflict without being censored. The watermelon emoji became a symbol for Palestine due to similar colors, replacing the Palestinian flag emoji after it was censored. These workarounds highlight how users adapt to arbitrary or discriminatory content policies.

The Impact on Marginalized Communities

Community guidelines, while aiming to prevent hate speech, often inadvertently silence marginalized communities. For example, Black creators using the n-word or queer creators discussing LGBTQ+ themes face removals or bans, as their language may overlap with “politically sensitive issues.” This creates a catch-22, forcing these groups to use algospeak like the ninja emoji for the n-word, or “zesty” for gay, hindering their ability to discuss their own experiences.

The Euphemism Treadmill and Algospeak

Steven Pinker’s euphemism treadmill describes the continuous cycle of words becoming negative and being replaced. Algospeak accelerates this treadmill by labeling words as undesirable, forcing rapid substitutions. This process can be seen in the shift from “suicide” to “unalive” and from “lesbian” to “le$bian” (which led to “le dollar bean” through TikTok’s text-to-speech). These changes occur rapidly and contribute to the evolution of language.

Algospeak as a Sociolect

Algospeak constitutes a sociolect—a form of language used by a particular social group. It serves as a common vocabulary reflecting shared online experiences and indicates belonging to the social media in-group. Like frat bro sociolects, it involves code-switching based on the domain of use. This linguistic adaptation allows users to express themselves effectively within the specific constraints of online environments, perpetuating itself through cross-platform content creation.

2: Sticking Out Your Gyat for the Rizzler

This chapter explores how social media accelerates language change through viral trends and algorithmic recommendations, leading to the rapid spread of new slang words. It highlights how the internet has democratized language, allowing non-standard forms to gain unprecedented popularity.

The Acceleration of Language Change

Historically, language change was slow due to insular communities and localized dialects. The centralization of England and the rise of print media began to accelerate this process, leading to the emergence of “slang” as a term for non-standard words. The internet, particularly social media platforms, has further intensified this acceleration, making language evolution faster than ever before.

Vine and the Birth of Viral Trends

The platform Vine, with its six-second video clips, was instrumental in popularizing viral trends and new slang words like “on fleek” and “yeet.” It also amplified existing terms from African American English like “bae” and “fam.” Vine’s content encouraged repetition and laid the groundwork for the more sophisticated recommendation systems of later platforms.

TikTok’s Algorithmic Amplification

TikTok perfected the recipe for addictive social media, utilizing a highly personalized algorithm that analyzes user behavior to recommend content. This algorithm amplifies trending songs and audios, creating a positive feedback loop where popular content becomes even more popular. The success of songs like Doja Cat’s “Say So” and Lil Nas X’s “Old Town Road” is directly attributable to their viral status on TikTok.

The “Side-Eye” Phenomenon

The “side-eye” meme is an example of a word gaining massive traction due to TikTok’s algorithm. Viral audio clips featuring phrases like “bombastic side-eye” led to a tenfold increase in Google searches for the term, demonstrating its significant impact on offline language use. TikTok’s personalized recommendations ensure that different users are exposed to different viral content, contributing to a diverse linguistic landscape.

Creators’ Strategic Use of Trends and Keywords

Content creators intentionally incorporate trending audios, words, captions, and hashtags into their videos to maximize algorithmic visibility. This strategic pandering, as highlighted by a University of Oxford study, means creators often minimize their own creativity to cater to algorithmic tastes. The engagement treadmill ensures that as certain content gains engagement, more creators produce similar content, further amplifying its spread.

The “Rizzler Song” and Gen Alpha Slang

The “Rizzler song”, with its slang-heavy lyrics (“gyat,” “rizzler,” “skibidi,” “fanum tax,” “sigma,” “ohio”), became a viral anthem for Gen Alpha humor. This song’s popularity drove the widespread adoption of these terms, with 80% of middle school teachers reporting hearing “rizz.” The Oxford English Dictionary even named “rizz” its 2023 Word of the Year, underscoring the profound impact of algorithmic trends on the mainstream lexicon.

Phrasal Templates and Language Play

The Rizzler song also highlights the role of phrasal templates (e.g., “on X,” “what the sigma”) in extending the lifespan of slang. These repeatable structures allow new words to be inserted into familiar linguistic formulas, contributing to language play and development. The word “function” becoming a synonym for “party” through meme templates like “White people when there’s X at the function” illustrates this process.

Memes as Viruses: Evolution and Selection

Memes are self-replicating units of culture that spread like viruses through social networks. In the algorithmic era, a “viral” video needs not just “likes” but also high watch time, comments, and shares. This new selective pressure means that words tied to easily replicable memes (like “sigma”) are more likely to survive and proliferate, often outcompeting potential synonyms due to their “funniness” and catchiness.

The Democratization of Slang

The internet has shattered traditional barriers to language dissemination. Content creators from diverse backgrounds now have a platform, allowing Black and queer slang to diffuse into the popular vernacular with unprecedented speed. This has transformed the definition of “slang” from a lower-class marker to a general term for informal speech used by everyone, reflecting the pervasive influence of the social media sociolect.

The Life Cycle of Memes: Adoption and Obtrusiveness

Everett Rogers’ Diffusion of Innovations theory categorizes adopters into innovators, early adopters, early majority, late majority, and laggards. The chair emoji () is an example of a meme that died out because its appeal was tied to an in-joke that lost value once it became mainstream and obtrusive. In contrast, the skull emoji () endured because it served a clear purpose in differentiating Gen Z from older generations.

The Endurance Factor and Semantic Gaps

For a word or meme to stick, it needs an endurance factor beyond initial virality. This includes appearing in many grammatical situations and filling a semantic gap—a necessary concept that previously lacked a word. Phrasal templates are crucial here, providing diverse contexts for words to be used. Words like “cancel” succeeded because they unobtrusively filled a semantic gap, while “yeet” and “on fleek” faded due to their obtrusiveness.

3: No Because What Happened to Your Attention?

This chapter focuses on how the attention economy drives linguistic evolution on social media. It explains how creators strategically use language and presentation to capture and retain user attention, leading to new communication patterns that often blur the lines between genuine interaction and manipulative tactics.

The Attention Economy and Virality

The attention economy, coined by Herbert A. Simon, views attention as a limited resource in an information-rich environment. On social media, creators are in a constant battle for user attention. Platforms like Reddit initially had transparent algorithms, allowing users to strategize posting times and titles to maximize upvotes. This demonstrated that even simple algorithms necessitate attention-grabbing tactics.

The Curiosity Gap and Emotional Language

To go viral, creators learned to exploit the curiosity gap—providing just enough information to intrigue users without revealing everything (e.g., “Snowfall in Sequoia National Park, California”). Emotional language also plays a crucial role; studies show that emotionally charged content is more likely to be shared. Creators use superlatives (e.g., “My favorite thing about X is…”) and extreme statements to hook viewers and elicit a strong response, driving engagement metrics.

Second-Person Pronouns and Parasocial Interactions

A common influencer tactic is the frequent use of second-person pronouns (“you,” “your”) to make content feel personally relevant and foster parasocial interactions. This frames the content in a way that viewers connect it to their own experiences, increasing watch time and engagement. Creators also leverage group dynamics by implying shared experiences (“Am I the only one who didn’t know X?”) to create a sense of inclusion and urgency to stay informed.

Ragebaiting and Clickbaiting

Some creators intentionally use incendiary or frustrating language to generate engagement, a strategy known as ragebaiting. Actress Louisa Melcher, for example, built a large audience by posting elaborate lies designed to spark strong emotional reactions. This strategy often blurs with clickbaiting, where enticing but sometimes deceptive promises are used to attract attention (e.g., “the eighth greatest fact of all time”). Paradoxically, “hate-watching” these videos still boosts engagement, perpetuating their spread.

Attention Retention Strategies in Short-Form Video

Social media platforms prioritize retention rate, the measurement of how long people watch videos. Influencers use various techniques to combat users’ short attention spans. These include microhooks within videos, deliberate scripting, and adapting content choices to trending topics. The MrBeast phenomenon exemplifies this, with his meticulous optimization of video pacing and extreme content to maximize engagement and virality.

The “Influencer Accent” as an Attention Tactic

New internet accents, particularly the “influencer accent” (often heard from female lifestyle creators), are fundamentally attention-holding strategies. Characteristics include uptalk (rising intonation at sentence end), emphatic prosody (stressing more words than necessary), and vowel lengthening. These techniques make sentences sound unfinished, inviting continued viewing, and create a conversational illusion that fosters parasocial connections and increases retention.

The “Millennial Pause” and “Gen Z Shake”

The stark contrast between how millennials and Gen Z start their videos highlights generational differences in attention tactics. The “millennial pause” (a brief silence at the beginning of a video) is derided by Gen Z, who immediately launch into talking. The “Gen Z shake” (a deliberate visual disruption from setting down the phone) is used to instantly capture attention and signal an interesting anecdote. These are conscious adaptations to the demand for constant stimulation.

Algorithmic Feedback Loops and Content Optimization

The pressure to gain attention creates a positive feedback loop: content that is slightly better at capturing attention performs exponentially better due to the Matthew effect and information cascades. This means successful linguistic techniques are self-perpetuated. Creators are incentivized to trendbait (coin new terms like “girl dinner” or “Roman Empire”) or use clever wordplay to generate comments and further boost their videos in the algorithm.

“No Because” as a Discourse Marker

The interjection “no because” has become a popular grammatical construction among younger generations. While seemingly unnecessary, it functions as a discourse marker to add emphasis, incredulity, or excitement. It serves as an attentional cue, signaling that engaging content is coming. This is a form of priming, where users are conditioned to associate the phrase with interesting videos, and it has seeped into offline conversation, demonstrating a direct influence of algorithmic tactics on everyday language.

Algorithmic Anxiety and Flanderization

The declining attention spans and pervasive algorithmic anxiety (a perceived lack of control online) are exacerbated by algorithms that prioritize engagement over quality. This leads to the proliferation of ragebait and misinformation, which capture attention even if they are detrimental. This also encourages Flanderization, where creators simplify and exaggerate their personas or content to fit what the algorithm rewards, leading to a shallowing of online identities and content.

4: Why Everybody Sounds the Same Online

This chapter delves into how social media algorithms contribute to the homogenization of language and accents globally, even while simultaneously fostering niche subcultures. It explores the pressures on creators to conform to an “Americanized” or “influencer” style of communication to maximize visibility and engagement.

The Americanization of Online Culture

Social media platforms, dominated by American tech companies, exert a powerful influence, leading to the Americanization of online communication. Many international creators feel pressured to “soften” their accents to sound more American, appealing to the large US audience. This phenomenon, noted in British children speaking with “YouTube accents,” contributes to the disappearance of regional accents worldwide.

The Rise of the “Influencer Accent”

A distinctive “influencer accent” has emerged online, particularly among female lifestyle creators. Key features include:

• Uptalk: Rising intonation at the end of sentences, making them sound like questions.
• Emphatic prosody: Stressing more words than necessary, drawing attention to keywords.
• Vowel lengthening: Dragging out final vowels.
• Increased rhoticity: Over-pronunciation of ‘r’ sounds.
  These elements are strategic, designed to maximize viewer retention by making sentences feel unfinished and engaging, thereby fostering a parasocial connection between creator and viewer.

Intentionality and the MrBeast Effect

Many creators consciously modify their speech to fit the online medium. MrBeast, a highly successful YouTuber, meticulously engineers his videos for maximum excitement and retention. His speaking speed has increased from 170 words per minute (wpm) to nearly 200 wpm, with reduced variance, demonstrating a deliberate optimization of his vocal delivery. This “entertainment influencer accent” is distinct from the “lifestyle influencer accent” but serves the same goal.

The “Educational Influencer Accent” and Behavioral Conditioning

Even educational creators like the author and Sophia Smith Galer adopt a unique “educational influencer accent,” characterized by faster, more energetic articulation and emphatic prosody, while retaining pitch variation. This is often a result of behavioral conditioning; creators observe what performs well and adjust their speaking style through trial and error, eventually becoming intuitive. This also includes trimming silences at the beginning of videos, leading to the “millennial pause” and “Gen Z shake” phenomena.

Subconscious Adoption and Prestige Dialects

Not all adoption of the influencer accent is conscious. People often imitate those around them, especially when there’s social value in sounding similar. This is seen in fraternities adopting a “fraccent.” The influencer accent functions as a prestige dialect, associated with the desirable social class of “influencers.” Viewers are primed to expect quality from this accent, leading them to engage more, further cementing its perceived value and widespread adoption.

Historical Roots of Media Accents

The influencer accent has historical roots. The “YouTube voice” (identified as early as 2015) shares features like overstressed words and lengthened vowels. This, in turn, draws from the Valley girl accent and the speaking styles of early celebrities like Paris Hilton and the Kardashians. The linguistic founder effect explains how the speech patterns of early, influential users become established norms, trickling down to micro-influencers and everyday users.

Diverse Yet Uniform: The Paradox of Online Accents

While there’s a trend towards online uniformity and Americanized accents, the internet also fosters diversity through specialized creator speaking styles. A beauty influencer’s nuanced approach differs from an educational creator’s fast-paced delivery. This paradox reflects social media’s ability to simultaneously homogenize and diversify language. Algorithms create a common online culture while segmenting users into niche groups that develop their own subcultures and speaking styles.

5: “The Algorithm Really Knows Me”

This chapter examines the paradoxical impact of algorithms on identity and language: while they promote cultural uniformity through viral trends, they also foster the growth of niche communities and their unique linguistic expressions. It reveals how algorithms “know” users and shape their online and offline personas.

The Rise of Fanilects and Niche Communities

Social media algorithms, by targeting content to specific interests, have driven a renaissance in niche communities and their specialized languages, known as fanilects. The “Swiftie fanilect” (Taylor Swift fans) and K-pop fanilect are prime examples, with their unique vocabulary, intertextual references, and portmanteaus. These fanilects allow in-groups to bond, share theories, and reinforce their group identity. The more specialized the language, the more precisely the algorithm can target content, deepening the filter bubble.

Homophily and Algorithmic Identity Construction

Humans have an innate tendency for homophily, “loving the same,” leading to the formation of in-groups based on shared interests. On social media, this translates into users gravitating towards specific content, which the algorithm then amplifies. Engagement metrics (likes, comments, shares) tell the algorithm whether to grant users deeper access to a group’s content. This creates a positive feedback loop where the algorithm plays a significant role in shaping our identities, as our perception of belonging is reinforced by personalized content.

Filter Bubbles and Echo Chambers

The concept of a filter bubble means algorithms prioritize content matching identified interests, leading to echo chambers that reinforce existing views. While often discussed negatively in terms of misinformation, every in-group on social media has its own filter bubble. For instance, K-pop stans are primarily recommended K-pop content, solidifying their identity and sense of belonging. The interaction between human behavior and algorithms is a complex, emergent system, where language and community shape each other.

The Emergent Effects of Human-Algorithm Interaction

The spread of fan vocabulary is an emergent effect of human-algorithm interaction. For example, the term “Gaylor” (a fan theory about Taylor Swift) emerged from a Swiftie in-group, and the algorithm then used it as metadata to further categorize and spread relevant content, reinforcing the Swiftie echo chamber. This dynamic applies to all new linguistic developments:

• Influencer accents and attention-grabbing phrases like “no because” emerge because they are effective at capturing and holding attention, which algorithms reward.
• Online fads and trending words like “rizz” and “skibidi” become amplified by algorithms, reaching broader audiences faster and influencing offline usage.
• Algospeak like “unalive” or “un@l!ve” is an emergent response to algorithmic censorship, creating new linguistic forms to circumvent restrictions.

Cyberbalkanization and Word Dissemination

The internet’s increasing division into special interest groups, known as cyberbalkanization, fuels the multiplication of new words. Most of these words remain within their niche in-groups, fulfilling specific semantic gaps (e.g., “Hiddleswift” for a celebrity couple). However, some words, like “in my X era” (from the Swiftie fanilect) or “delulu” (from K-pop culture), escape their original communities and enter mainstream usage through broader memes and trends, amplified by the algorithm’s ability to push content across fuzzy filter bubble boundaries.

Niche Communities as Early Adopters

Niche communities act as “early adopters” of new words, creating specialized terminology for their shared experiences or innovations (e.g., tech nerds coining “software,” “hyperlink”). While many terms remain niche (e.g., anime tropes like “yuri”), some spread when tied to important cultural concepts, often as memes. “Sussy baka,” an anime-derived term, spread as a humorous expression beyond its original in-group, demonstrating how algorithms accelerate the diffusion of culturally valuable words.

Context Collapse and Misinformation

The fuzzy boundaries of filter bubbles lead to context collapse, where communication intended for a niche audience reaches unintended audiences that may interpret it differently. This is problematic for terms like “acoustic” (for “autistic”) or “neurospicy” that originated as in-jokes within neurodivergent communities but were later used negatively by outsiders. Algorithms exacerbate this by pushing content that sparks division and misinformation, as engagement (even negative comments) boosts visibility. This includes autism misinformation on TikTok, leading to self-diagnosis based on inaccurate information and the misuse of in-group language.

Engagement Optimization and “Digital Rubbernecking”

Engagement optimization algorithms, designed to maximize user interaction, prioritize content that generates comments, shares, likes, and high retention. This means that “fun” or “entertaining” content, including trending slang, performs better, encouraging its use. However, this also rewards “digital rubbernecking”—people’s tendency to react to negative or annoying content. This leads to the proliferation of ragebait and incivility, even if users don’t consciously want to consume such content, creating a system where “Goodhart’s law” applies: optimizing for engagement leads to unwanted outcomes.

6: Wordpilled Slangmaxxing

This chapter delves into how extreme online communities, particularly incels, generate highly specific slang that, through the power of memes and algorithms, can infiltrate mainstream language. It explores the blurred line between irony and authenticity, and the potential dangers of fringe ideologies spreading through seemingly innocuous humor.

Incel Slang: A Case Study in Extreme Subcultures

The incel (involuntary celibate) community developed a detailed philosophy and specific slang to describe their worldview, including terms like “blackpilled” (acceptance of lookism), “Chads” (attractive men), “betas” (average men), and “looksmaxxing” (enhancing physical appearance). This highly specialized vocabulary initially served as an in-group marker within their forums (e.g., 4chan and Reddit subreddits).

The Cult-like Nature of Incel Language

Incel language functions similarly to cult language, serving as a recruitment tool that creates an “us versus them” mentality. Terms like “mogging” (dominating) and “cucked” (emasculating) reinforce a reductive worldview, where social interactions are framed hierarchically and women are depersonalized as “foids” or “dumpsters.” This linguistic immersion can indoctrinate members, limiting their introspection and reinforcing extreme beliefs like male supremacy and the justification of sexual violence.

From Fringe Forums to Mainstream Memes

Despite the extreme nature of incel ideology, their slang and concepts have diffused into the mainstream, primarily through memes and algorithmic amplification. Initially, incel jargon spread from 4chan to less extreme Reddit forums like “rate me” subreddits, where looksmaxxing language was disguised as helpful beauty suggestions. Women on #GirlTok unknowingly adopted and repurposed these concepts, using terms like “canthal tilt” and “mewing” in beauty tutorials, which then spurred further memetic spread and led to the widespread adoption of suffixes like “-maxxing” and “-pilled.”

The Algorithmic Amplification of Incel Memes

Once these incel terms captured mainstream curiosity, the Matthew effect and engagement treadmill amplified their spread. Their memetic value and recombinability into phrasal templates (e.g., “what the sigma”) ensured their virality. Algorithms, designed to maximize engagement, pushed content containing these words, further popularizing them. Even when the original incel communities moved to private, harder-to-find forums, their language continued to proliferate on mainstream platforms.

The Dangers of Unwitting Appropriation

The widespread use of incel memes, like “Chad walk versus virgin stride” or “Gigachad,” often occurs without knowledge of their origins, leading to the unwitting perpetuation of harmful ideas. Concepts like lookism and its associated pseudoscientific beauty standards (e.g., interocular distance, hunter eyes) have entered mainstream discourse. The “Oxford study” meme, referencing a fictional paper on WMAF (white male/Asian female) relationships, demonstrates how misogynistic scrutiny rooted in incel beliefs can become a widespread form of online harassment.

Doomslang and Incel Influence

The rise of “doomslang” (e.g., “it’s over,” “bedrotting,” “wagecuck,” “brainrot”) reflects a growing pessimism among younger generations and shows influence from incel vocabulary. Terms like “LDAR” (“lay down and rot”) and “doomer” originated in incel circles. This language spreads because algorithms thrive on negativity and such phrases confirm existing cultural outlooks, creating a self-reinforcing cycle where apocalyptic statements become part of the zeitgeist.

Irony, Authenticity, and Poe’s Law

The spread of incel memes highlights the blurred line between comedy and authenticity on the internet. Many users share these memes purely for humor, satirizing the original incel ideology. However, Poe’s law explains how ironic expressions of extreme views can be mistaken for sincere ones, and vice versa. This means “edgy” humor can inadvertently normalize dangerous ideas. When some beauty influencers treated lookism terms seriously, it led to a circular effect where jokes inspired genuine belief, which then spurred more jokes, furthering the concepts’ cultural relevance.

The 4chan Legacy and Meme Marketing

The diffusion of words like “seggs” and “unalive” from 4chan and Reddit to TikTok demonstrates that niche communities (often those with a “need to invent slang”) are linguistic innovators. Research suggests that a plurality of internet memes originate from political forums like 4chan, often used as a form of “attention hacking” to spread ideologies. In the modern attention economy, companies like Duolingo also leverage meme marketing (e.g., posting its mascot mewing) to connect with younger audiences, showing how even extreme ideologies can diffuse through humor.

7: It’s Giving Appropriation

This chapter explores how African American English (AAE) and its subcultures consistently serve as sources of linguistic innovation, with their terms often being appropriated by mainstream internet culture. It examines the process of linguistic gentrification, where words lose their original meaning and power as they spread through broader social networks.

The Cycle of Linguistic Appropriation: “Cool”

The word “cool” exemplifies linguistic appropriation. Originating in African American English (AAE), it conveyed a demeanor of calmness and resistance. Its alluring exclusivity led to its adoption by counterculture communities (jazz, punks, beatniks) and then the mainstream, losing its original nuance and cultural value to the Black community. This constant need for new words, due to previous terms being appropriated, fuels the creation of new AAE slang like “hip” and “woke.”

AAE as a Source of Online Slang

AAE serves as a major source for “internet slang.” Words like “bae,” “fleek,” “fam,” “cap,” “sis,” “bruh,” and “bussin” all have AAE origins. Grammatical constructions like “not you,” “it’s giving,” and “the way you X” are also derived from AAE. These terms become popular in mainstream online culture because they are perceived as “cool” or countercultural, perpetuating a cycle of appropriation where their origins are often forgotten.

Ball Culture and Its Linguistic Legacy

Ball culture, an underground LGBTQ+ movement predominantly among Black and Latino communities in New York City, developed a rich, specialized vocabulary. Terms like “throw shade,” “give cunt,” “slay,” “work,” “serve,” “eat it,” “tea,” “gagging,” “icons,” “queens,” and “yass” all originated in this defiant subculture. Mass media (e.g., RuPaul’s Drag Race, Beyoncé’s Renaissance) played a crucial role in popularizing this argot to wider audiences, often without proper attribution or understanding of its original context.

The Diffusion of Ballroom Slang

The spread of ballroom slang follows a predictable pattern of diffusion from niche to mainstream. It started with a highly marginalized “innovator” group (ball houses), moved to broader gay communities (perceived as “cool”), and then diffused to young women and general internet users (perceived as “cool” by them). This process, amplified by social media, leads to context collapse, where words are used outside their original meaning and often become awkward or offensive to their originators (e.g., “slay” losing its meaning for queer Black people).

“Hood Irony” and Gang Slang

“Hood irony” is another genre of online humor that appropriates language from gang culture. The Blood street gang’s practice of replacing ‘c’ with the ️ emoji evolved into ironic online usage, leading to its misuse as a substitute for the n-word. Similarly, Crip gang habits (replacing ‘ck’ with ‘cc’ in words like “thicc,” “succ,” “fucc”) became mainstream memes. The word “opp” (from Chicago gang scene) also became widely accepted as “Gen Z slang” without knowledge of its violent origins, trivializing serious issues.

The Problematic Humor of AAE Pronunciations

Many online appropriations of AAE are rooted in the problematic perception of Black culture as “funny.” Exaggerated AAE pronunciations, like “gyat” (from “goddamn”) or “ahh” (from “ass”), become “internet brainrot slang,” further diluting their original meaning and contributing to false etymologies. This unintentional reinforcement of racist stereotypes (e.g., Black people as overly expressive) perpetuates negative generalizations.

Digital Blackface and “Blaccents”

Digital blackface refers to the disproportionate use of reaction GIFs and images of Black people by non-Black individuals, subtly playing into racial stereotypes. The emergence of “Blaccents“—fake AAE accents used by non-Black individuals (e.g., Tray Soe’s “cone bread”)—highlights how non-Black creators mimic Black speech to tap into its perceived coolness. This “slippery slope” normalizes the adoption and caricaturing of African American culture, often without the original communities benefiting or receiving credit.

The Loss of Ownership and Value

The rapid, trend-based nature of memes on social media directly causes a loss of ownership for linguistic innovators. Kayla Newman, who coined “fleek,” never received royalties despite celebrities and corporations profiting from the term. Jalaiah Harmon, who choreographed the “Renegade” dance, received no credit while Charli D’Amelio gained millions. This illustrates how larger accounts benefit from smaller accounts, often without attribution, stripping words and creative works of their original power and significance.

8: What Are We Wearing This Summer?

This chapter explores how algorithms, particularly through search engine optimization (SEO) and engagement optimization, drive the creation of hyper-specific fashion aesthetics and microlabels. It reveals how social media platforms and businesses co-opt language to commoditize identity and leverage the long-tail economic model.

Keywords and Algorithmic Optimization

Keywords are crucial for algorithmic optimization on social media, going beyond traditional search engine optimization (SEO). Platforms like TikTok reward creators for including relevant titles, hashtags, and captions, as this provides metadata for their algorithms to categorize and recommend personalized content. This incentivizes creators to spread or create labels that algorithms can easily process.

The Proliferation of “-core” Aesthetics

The suffix “-core” (e.g., “cottagecore,” “Barbiecore”) is a prime example of a fashion microlabel denoting a specific aesthetic. The explosion of these “-core” terms, particularly since 2020 and the rise of short-form video, serves as a highly specific way to inform algorithms about user preferences. Each interaction with a “goblincore” video, for example, feeds the algorithm data, leading to more personalized recommendations and pushing users deeper into specific filter bubbles.

Microlabels Across Aesthetics and Music

The phenomenon of microlabels extends beyond fashion (“coquette,” “dark academia”) to music genres. Spotify, for instance, has created over six thousand “microgenres” (e.g., “bedroom pop,” “hyperpop”) to better categorize music. These labels, even if invented by the platform, influence how humans talk about and identify with these genres. The more specific the labels, the more granular the personalized recommendations, allowing platforms to cater to increasingly niche interests.

Identities as Commodifiable Demographics

These new microlabels package entire lifestyles, enabling users to delineate their identities more meticulously than ever before (e.g., “vampire goth” vs. “pastel goth”). Social media enables the discovery of these niche communities, which then form around new aesthetics and coin new words. This creates new “subcultural demographics” that businesses can market to, turning identity into a commodifiable asset. Platforms actively incentivize this through integrated online shops and influencer marketing, where creators promote products to their niche communities.

The Long-Tail Economic Model

The hyper-compartmentalization of cultural labels aligns with the long-tail economic model, where companies sell small quantities of unique items to niche audiences. Platforms like TikTok, Instagram, and YouTube leverage their vast user data to identify and serve these niche “demand tails.” They encourage creators to specialize in these niches, or to jump on trending “–core” aesthetics, reinforcing the idea that identity is interchangeable with metadata, which is then used for profit.

Enshittification and Language

Cory Doctorow’s “enshittification” describes the inevitable decline in quality on social media platforms as they prioritize profit over user experience. This manifests through algorithms that increasingly push less organic, engagement-maximizing content (e.g., more ads, trending memes) rather than content aligning with users’ conscious preferences. The proliferation of new microlabels and algorithmic spamdexing (e.g., excessive hashtags) is a direct consequence, as creators and businesses desperately try to gain visibility in this degraded environment.

Linguistic Compression and Loss of Nuance

Paradoxically, the influx of new microlabels can lead to linguistic compression and a lack of nuance. While categories become more specific, the terms themselves often remain formulaic (“adjective + core”). Creators may prioritize trending, generic terms over precise vocabulary to maximize virality, even if less accurate. This can lead to semantic drift, where words like “preppy” dramatically shift meaning to align with commercial trends. This creates a system where identity is shaped by predefined boxes, potentially flattening true individuality in favor of what algorithms prefer.

9: OK Boomer

This chapter explores how the concept of “generations” has evolved and been amplified by social media, leading to intergenerational tribalism and the creation of new “generational slang.” It argues that while these labels are often arbitrary, they hold significant cultural power due to algorithmic reinforcement and the human tendency to categorize.

The Evolution of Generational Labels

The modern idea of a “generation” (e.g., “millennials,” “baby boomers”) emerged in the 20th century due to shared historical experiences. These labels, though often arbitrary and lacking scientific backing, have gained immense cultural salience in the social media era. Algorithms contribute to this by categorizing users into these cohorts, driving personalized recommendations and creating new demographics for marketing.

Generational Cultural Polarization

Different generations exhibit subtle cultural polarization, particularly in humor and technology use. Boomers are caricatured for their “newspaper cartoon” humor, millennials for “LOLcats” and “image macros,” and Gen Z for “wry yet detached absurdism.” Even emoji usage differs, with younger generations increasingly abstracting meanings (e.g., skull emoji for humor). These perceived differences are amplified by memes, fueling intergenerational “wars” like “OK boomer” or “millennial pause” jokes.

The Rise of “Generational Slang”

The increased importance of generational labels has led to the notion of “generational slang” (e.g., “mewing,” “cap,” “ate” attributed to Gen Z). While slang has always existed, social media’s rapid trends and hard-to-trace diffusion allow words to be quickly associated with specific generations. This makes the language obtrusive, facilitating its dismissal as “improper English” and contributing to generational friction.

“Brainrot” and the Stigmatization of Slang

The term “brainrot” emerged to describe Gen Alpha slang (e.g., “rizz,” “gyat,” “skibidi”), implying it’s indicative of unhealthy online presence and a decline in intelligence. This stigmatizes normal etymological processes and harms the origin communities (e.g., AAE speakers). Media attention, like the New York Times article on “Gen Alpha” slang, further reinforces these generalizations, creating a positive feedback loop where curiosity about “generational slang” drives its virality.

Flanderization of Online Identities

Flanderization—the simplification and exaggeration of characters—is rampant on social media. Creators hyperbolize their personas to optimize for engagement, leading to typecasting. This applies to word choice: if “Gen Z” yields more views than “young people,” the simpler label proliferates. This also results in the Flanderization of subcultures (e.g., “preppy” becoming a caricature of the popular-girl aesthetic), as extreme representations are amplified by algorithms, shaping perceptions and normalizing those extremes.

Social Learning and Algorithmic Influence

Algorithms subtly influence social learning by presenting users with the most extreme behaviors of others, leading to an overestimation of differences and a tendency to imitate these extremes. This can lead to conscious adoption of “generational behaviors” (e.g., Gen Z finger heart). Generational labels, like other microlabels, act as stories and categories that define our identities, but this can constrain self-expression by fitting individuality into predefined boxes.

Internet Folklore and Lore

Social media creates a shared canon of references that forms internet folklore. Influential videos and memes (e.g., “Bella Poarch head-bopping video,” “Gen Z intern”) are constantly remixed and referred to, becoming part of a subculture’s shared history. The word “lore” itself has seen a fourfold increase in Google searches, with younger people applying it to their own pasts or mundane events (“Apricot lore is crazy”), reflecting a reflexive awareness of this narrative-building process.

10: Are We Cooked?

This final chapter synthesizes the book’s arguments, asserting that while social media introduces new linguistic challenges, humanity’s innate adaptability ensures that language will continue to evolve. It addresses the pervasive “cooked” (screwed) sentiment about language decline, arguing for an optimistic view rooted in historical linguistic patterns.

The Globalization of Slang

Social media has made slang unprecedentedly global. English words and phrases (e.g., “delulu,” “mewing,” “slay,” “serving”) are rapidly adopted and calqued into other languages like Spanish, Portuguese, and Arabic. This demonstrates the increased impact of English as the lingua franca online, where trends quickly cross linguistic boundaries, often losing their original context in the process.

Decentralization of Language Control

The internet has decentralized language control, diminishing the influence of traditional language planning institutes like the Real Academia Española (RAE) and the Académie Française. Users, particularly younger generations and those in former colonies, are increasingly innovating language independently, incorporating loanwords and nonstandard forms that prioritize natural usage over institutional norms. This is a form of decolonization of language, giving everyday citizens more power over communication.

The Unwinnable Whac-A-Mole Game

Censors cannot win the Whac-A-Mole game against linguistic creativity. Even in highly censored environments like China, users invent clever algospeak substitutions (e.g., “héxié” for “harmonious,” “river crab” for “censorship”), making comprehensive censorship practically impossible without disabling platforms. The Kongish language used during the Hong Kong protests exemplifies how playful, meme-based language can defy government inhibition and build community.

Politics and Media Adaptation

Politicians like Senator Fatima Payman and Senator Bernadette Clement are increasingly incorporating online slang (“sigmas,” “goofy ahh,” “capping,” “yikes”) into their speeches for shock value and to gain attention, showing how governmental institutions adapt to the algorithmic imperative for relevance. Traditional news outlets struggle to compete with social media creators like Dylan Page (“News Daddy of TikTok”), who prioritize engagement tactics over traditional journalistic standards, often leading to sensationalism and misinformation.

The Paradox of Linguistic Diversity

The impact on linguistic diversity is paradoxical. While English benefits from preferential attachment and regional dialects within English (e.g., Texas accent) are declining, social media also fosters new specialized dialects within online filter bubbles (e.g., Swiftie dialect, incel dialect, furry dialect). This suggests a shift from geographic to digital linguistic diversity, with individuals developing multiple “accents” as they code-switch between online and offline contexts.

ASL as a Case Study of Adaptation

American Sign Language (ASL) provides a visible example of how language adapts to social media constraints. Signs for words like “dog” have changed to accommodate vertical video (e.g., signing DG twice). Deaf users often adapt to one-handed signs and tighter signing spaces. While social media connects the geographically dispersed Deaf community, it also creates issues like hearing ASL creators gaining disproportionate visibility and the spread of inaccurate ASL “gibberish”, diluting the language’s original importance.

Optimism in the Face of Algorithmic Anxiety

The widespread “algorithmic anxiety” reflects concerns about censorship, surveillance, and deindividualization. However, Aleksic argues for an optimistic perspective. The new ASL sign for “dog” is not societal collapse but human adaptation. Language change has always been influenced by invisible factors and reflects cultural moments. The book’s title “Algospeak” initially referred to censorship avoidance, but its scope broadened to encompass how every aspect of language is shaped by algorithms.

Humanity’s Enduring Adaptability

Drawing on Geoffrey Chaucer’s observations about language change, Aleksic concludes that despite constant linguistic evolution, human nature endures. “Algospeak” is not a story about algorithms dominating humans, but about humans adapting to algorithms. We remain creative, resilient, and ingenious in our use of language, constantly finding new ways to express ourselves, form communities, and navigate our reality, just as we have through every technological and cultural revolution in history.

Key Takeaways: What You Need to Remember

Core Insights from Algospeak

• Algospeak is the new reality of language, driven by social media algorithms that shape how words spread, are used, and evolve.
• Censorship workarounds are a primary driver of algospeak, creating a continuous Whac-A-Mole game between users and platforms.
• Viral trends, memes, and phrasal templates are key mechanisms for language diffusion, accelerating the spread of new terms.
• Attention economy tactics, such as emotional language, superlatives, and specific vocal patterns (influencer accents), are intentionally used by creators to maximize engagement.
• Niche communities and fanilects thrive in filter bubbles created by algorithms, fostering unique vocabularies that can sometimes bleed into mainstream usage.
• Linguistic appropriation, particularly of African American English and LGBTQ+ slang, is rampant, leading to words losing their original context and power.
• Generational labels and their associated “slang” are amplified by algorithms, contributing to intergenerational tribalism and oversimplified identities.
• Microlabels and aesthetics are a retail strategy to commoditize identity, aligning with the long-tail economic model and shaping consumer preferences.
• Algorithms don’t truly “know” you; they predict engagement based on past behavior and then subtly guide your identity through content recommendations.
• Language change, though now accelerated and influenced by technology, is fundamentally a human process of adaptation and creativity.

Immediate Actions to Take Today

• Observe your own language use on social media and in daily conversation to identify instances of algospeak and algorithmic influence.
• Critically evaluate trending content to discern its origins, potential biases, and whether it’s genuine or a form of ragebait/clickbait.
• Be mindful of the words you adopt from online trends, especially those originating from marginalized communities, and consider their original context.
• Diversify your media consumption to break out of algorithmic filter bubbles and expose yourself to different perspectives and linguistic styles.
• Engage thoughtfully with online content, knowing that your likes, comments, and shares contribute to the algorithmic feedback loops.

Questions for Personal Application

• How do the algorithmic pressures described in the book manifest in your daily social media use (e.g., why do certain types of videos consistently appear on your feed)?
• In what ways have your own speech patterns or vocabulary changed due to your time spent online, and do you consciously or subconsciously adopt “influencer accents” or slang?
• How do you perceive the authenticity of content creators online, and how might your perceptions be influenced by their use of attention-grabbing language or exaggerated personas?
• When you see new slang or memes, do you consider their origin and the communities from which they emerged, or do you primarily focus on their humorous or trending aspect?
• How might understanding the commodification of identity and the long-tail economic model change your approach to online shopping or personal branding?
• What actions can you take to “train your algorithm” to better align with your stated preferences rather than just your revealed, emotional reactions?
• How does the concept of linguistic appropriation resonate with your understanding of cultural exchange, and what role do you play in this process?
• Reflect on how generational labels influence your interactions and perceptions of others; can you challenge these Flanderized stereotypes in your conversations?
• Considering the ongoing evolution of language, what responsibilities do you believe individuals and platforms have in shaping the future of communication?