Gender coding is used when assigning specific traits or behaviors primarily or exclusively to certain genders. Gender codes are often presented by words and phrases associated with a particular gender, specifically male or female.
These words and phrases reflect a culture’s ideas about gender and are often based on stereotypes.
Examples of typically ‘masculine-coded’ words that might be used include terms such as competitive, dominate, driven, and challenge. Examples of typically ‘feminine-coded’ words include loyal, collaborative, understanding, and encouraging.
Key Takeaways
- Definition: Gender coding assigns specific traits or behaviors mainly to one gender, based on cultural stereotype rather than fact.
- In Job Ads: Masculine-coded words such as ‘dominant’ and ‘competitive’ appear often in postings and can discourage women from applying, even when they are qualified.
- In Evaluations: Communal, feminine-coded praise (e.g., ‘helpful’) is common for women but rarely rewarded; agentic, masculine-coded language (e.g., ‘takes charge’) is.
- Theory: Goffman, West and Zimmerman, and Connell each explain gender coding as a learned, socially enforced sign system, not a fixed trait.
- Limits: Much of the evidence is workplace-specific, correlational, and still frames gender as a strict binary.
Where Is Gender Coding Used?
Gender coding shows up most clearly at work: in job advertisements, in performance evaluations, and even in how violence gets judged.
Gender Coding in Job Advertisements
Job advertisements are a well-studied site for gender coding. Employers often list personal attributes, such as motivation and drive, rather than factual requirements like experience and qualifications. These words are themselves gender-coded. So a posting can end up signalling, often unintentionally, which gender is expected to apply.
Aim: To test whether finance internship postings use masculine-coded language, and whether this affects how suitable applicants judge themselves to be for the role (Oldford & Fiset, 2021).
Method: A content analysis coded masculine-coded terms, like dominant and competitive, in finance internship postings. A linked survey measured applicant fit ratings too.
Results: Masculine-coded words appeared often, usually framed as essential qualities. Female respondents rated themselves as less suitable for masculine-coded postings than male respondents did, even when requirements were identical.
Conclusion: Finance recruitment language is not gender-neutral. This imbalance appears to shape who applies, separate from applicants’ actual qualifications.
A similar pattern appears in emergency medicine. These job postings tend to use more masculine-coded language. In one study, the large majority of advertisements contained at least one gender-coded word, and masculine-coded language was especially concentrated in leadership positions (O’Brien et al., 2022).
The jobs themselves can be gender-coded too. This happens independent of how a posting is worded (Grip & Jansson, 2022).
Occupations and work tasks are often assumed to suit one gender more than another (Acker, 1990). The labor market itself helps create and reinforce these gendered identities (McDowell, 2004).
Gender Coding in Performance Evaluations
Gender codes don’t only shape who applies for a role. They also shape how an employee’s performance is described and rewarded once they are in it.
Aim: To determine whether performance evaluations use gendered language differently depending on an employee’s gender, and whether that difference affects how favorably the person is rewarded (Correll et al., 2020).
Method: Researchers analyzed real evaluations.
They coded language for communal terms (e.g., helpful, supportive, collaborative) and agentic terms (e.g., confident, decisive, “takes charge”), then cross-referenced this against each employee’s gender and career outcomes.
Results: When communal language appeared in an evaluation, it was applied to women 61% of the time. Being called “helpful” did not translate into better ratings. Employees described in agentic terms, disproportionately men, were rewarded for it.
Conclusion: The organization’s reward structure was not gender-neutral. Agentic, masculine-coded behavior was valued; communal, feminine-coded behavior was not, even when it was genuinely positive.
Gender Coding and the Categorization of Violence
Gender codes even shape how identical actions are judged. Physical violence tends to be coded as masculine, and relational violence, such as social exclusion or rumor-spreading, as feminine.
Boys are often described as less violent when they use relational aggression. Physical violence by girls is not always taken as seriously (Lunneblad & Johansson, 2021).
This works like a shifting threshold. The point at which a behavior is recognized and responded to as a problem moves, depending on whether it matches or breaks the gender code expected of the child involved.
A boy’s relational aggression falls outside the masculine script, so it gets minimized. A girl’s physical aggression falls outside the feminine script, so it gets minimized too, just in the opposite direction.
How Gender Codes Work
Sociologists have asked why gender codes feel natural rather than arbitrary. Three accounts stand out.
Goffman’s Visual Grammar of Gender
The sociologist Erving Goffman produced one of the first studies of gender coding. He analyzed how men and women were posed in advertising images, not in language (Goffman, 1976).
He found a consistent set of visual codes. Women were often shown gazing away from the camera, lying down, or touching objects gently rather than grasping them. Men stood taller. They were posed as instructors, gripping objects firmly.
Goffman called this pattern hyper-ritualization: advertising exaggerates ordinary gestures of dominance and deference into an easily readable shorthand. Later studies applying his categories to modern advertising have found the same patterns, still present but somewhat softened.
The method has a limit, though. Goffman’s categories came from his own reading of a small, non-random image sample, covering only print advertising in one culture and era.
Goffman’s wider theory of everyday self-presentation is explored further in impression management.
Doing Gender: A Social Accomplishment
The sociologists Candace West and Don Zimmerman argued gender is not a fixed trait. It is something people actively do (West & Zimmerman, 1987).
A society first sorts people into gender categories. Individuals are then held accountable, in everyday interaction, for behaving in ways that match their category.
A man in a suit is read as ordinary. A woman in the identical suit is more likely read as making a statement. Both are ‘doing gender’: performing a legible, gender-coded act that other people interpret through the lens of gender.
A code, on this account, is not just a statistical pattern, like ‘competitive job-ad language happens to attract men.’ It is a standard people are held accountable to.
This explains why codes feel obligatory, not optional. A departure from the standard needs to be explained. Conforming to it needs no explanation at all.
Hegemonic Masculinity and Emphasized Femininity
Gender codes are not equal opposites, either. The sociologist Raewyn Connell, writing later with James Messerschmidt, argued that every culture holds up one dominant, idealized form of masculinity as its standard (Connell & Messerschmidt, 2005).
This is called hegemonic masculinity: the standard against which other masculinities and femininities are measured and ranked.
It legitimizes men’s dominance over women. It also ranks masculinities against each other. Those coded as insufficiently masculine are subordinated.
The matching concept, emphasized femininity, describes femininity organized around accommodating this dominant masculinity, through compliance, nurturance, and attractiveness, rather than around women’s own authority.
Feminist sociologists (see feminist theory) argue this hierarchy is why gender coding is not neutral. Masculine-coded traits align with traits rewarded in high-status roles, while feminine-coded traits align with traits expected in supporting ones.
These three accounts explain how codes operate day to day. They are distinct from the broader process of gender socialisation, through which people learn a culture’s gender norms generally, from family, peers, and media.
What Is The Purpose Of Gender Coding?
Gender codes are rarely used on purpose. But they still send a subtle message about what is acceptable for men and women.
Take gender-coded violence as an example. Boys and men may be excused when they show relational forms of violence. Girls and women, by contrast, often learn that displaying the same behavior is less acceptable for them.
Any gender can show masculine- or feminine-coded traits. But coding still serves a purpose in job advertisements.
Heavily masculine-coded language can signal that a company is mostly staffed by men. It can suggest, even unintentionally, that other genders are not especially welcome there.
Feminine-coded language can do the reverse. It can make a job feel unappealing to certain genders, whether or not that is the employer’s actual intention.
What Is The Impact Of Gender Coding?
Gender coding keeps stereotypical ideas about each gender alive. This happens even when the ideas are not true of any individual.
In many countries, even ones seen as gender-progressive, men are still perceived as more agentic, while women are seen as more communal (Oldford & Fiset, 2021).
To be agentic means to be confident and decisive. These are qualities typically ascribed to men. To be communal means to be warm and supportive. These qualities are typically ascribed to women.
Because of this, women may feel less comfortable applying for jobs that seem to be seeking male candidates. They may apply instead for postings using terms such as ‘interpersonal’ and ‘caring.’
Both men and women can lose out here. A posting’s language may simply not feel like it applies to their gender.
This kind of self-selection can segregate workplaces by gender. If mostly women apply to feminine-coded postings and mostly men apply to masculine-coded ones, occupations stay divided along gender lines (Grip & Jansson, 2022).
This same pattern shows up in performance evaluations. Positive, feminine-coded praise, such as being called ‘helpful,’ is not rewarded the way agentic, masculine-coded language is. This has been linked to real differences in pay and promotion (see Gender Coding in Performance Evaluations, above; Correll et al., 2020).
How To Avoid Gender Codes In A Job Description
Using non-gendered language in job descriptions and advertisements encourages people of all genders to apply. It also supports a more diverse and inclusive workplace.
Using a balance of masculine- and feminine-coded terms, or avoiding gendered language altogether, helps too. It stops a posting from subtly signalling who is expected to apply.
Superlatives like ‘superior’ and ‘expert’ skew masculine. They can discourage women, who are on average more inclined toward collaboration than competition, from applying. Gender-neutral job titles help too, and can attract people of every gender to apply.
Critical Evaluation
The evidence for gender coding is strong in places and thinner in others. Five issues are worth weighing.
- Ecological Validity: the workplace studies use real organizational data, but each comes from a single country, industry, or organization.
- Cultural Specificity: what counts as ‘masculine-coded’ or ‘feminine-coded’ is a historically and culturally specific judgement, not a fixed fact.
- Correlational Evidence: most studies show coded language and unequal outcomes moving together, without proving what causes what.
- A Persistent Binary: almost all of the evidence sorts people into exactly one of two coded categories.
- Contemporary Research: a 2023 survey experiment offers the field’s clearest causal evidence on how unevenly codes are enforced.
Ecological Validity of the Workplace Evidence
Several of the studies above analyzed real organizational data, not laboratory vignettes. Correll et al. (2020) coded actual performance reviews used for real promotion decisions. This is real data, not a hypothetical.
Oldford and Fiset (2021) paired a job-posting content analysis with a matched applicant survey. Together, these give the findings real practical weight.
But the evidence is narrow. Most workplace studies come from a single country, industry, or organization. Examples include a Swedish school system, one American technology company, and the finance and emergency-medicine sectors. An exact figure, like the 61%, may not generalize elsewhere.
That specific number should not be assumed to hold in other industries or countries, even if the general pattern likely does. The direction, though, seems robust.
A Culturally and Historically Specific Coding Scheme
Calling a word ‘masculine-coded’ or ‘feminine-coded’ depends on a cultural judgement, not a fixed fact. The pink-for-girls, blue-for-boys convention feels timeless. Historical research on American children’s clothing found the opposite (Paoletti, 2012).
Early twentieth-century trade advice often recommended pink for boys and blue for girls. Pink was reasoned to be the stronger, more ‘decided’ colour.
The now-familiar reversal only settled around the mid-twentieth century, and hardened further from the 1980s onward. A code that feels obvious today was, within living memory, coded the opposite way.
The same problem runs deeper than one colour convention. Word lists used to code language as masculine or feminine were built and tested on contemporary, Western, English-language samples. Applying them elsewhere risks assuming a word’s coding transfers unchanged across cultures.
The labels can mislead. A study using different or dated word lists may not even be measuring the same underlying construct, despite using identical labels.
Correlational, Not Causal, Evidence
Much of the evidence shows that coded language and unequal outcomes occur together. It does not always prove what causes what. Correll et al. (2020) and Grip and Jansson (2022) each document an association. Neither randomly assigned the language people actually used.
A correlational study can show that gender-coded language and career outcomes move together.
It cannot show, on its own, that changing the language would change the outcome. That distinction matters.
Anyone designing a fix, such as rewriting a job advertisement or retraining evaluators, needs more than an association to know it will work. Khanna and Meadow’s (2023) experimental design, discussed below, is a notable exception, because it manipulates the child’s described sex and behavior directly, allowing a genuinely causal reading.
A Persistent Masculine/Feminine Binary
Most of the workplace evidence, including Oldford and Fiset (2021) and Correll et al. (2020), sorts people into exactly one of two coded categories. The question is usually how the binary gets enforced.
It rarely asks whether the binary is the right frame at all. Very little of the literature measures how gender coding works for people outside that binary.
This is a real gap. One open question is whether the word-coding lists used throughout this research even apply to non-binary or gender-nonconforming language. No study reviewed here has tested this directly.
That gap matters for practice, not just theory. A word-coding scheme validated only on a strict binary sample cannot be assumed to classify non-binary or gender-nonconforming language the same way. No study in this literature has checked.
Contemporary Research
The clearest recent test of how unevenly codes get enforced comes from a 2023 survey experiment.
Aim: To test whether a child’s assigned sex affects how readily adults reclassify that child as transgender, when the child’s behavior breaks their expected gender code (Khanna & Meadow, 2023).
Method: A survey experiment used a nationally representative U.S. sample.
Respondents read vignettes describing a child showing gender-nonconforming behavior, a boy acting feminine-coded or a girl acting masculine-coded.
They then rated how likely they were to reclassify the child as transgender.
Results: Boys were reclassified as transgender far more often. This happened when boys showed feminine-coded behavior, more than when girls showed equivalent masculine-coded behavior. Membership in the ‘male’ category proved more fragile than membership in the ‘female’ category.
Conclusion: Gender nonconformity is not symmetrical. Because the design was experimental and nationally representative, this offers some of the strongest available evidence that gender codes are enforced unevenly. The durability of a social category depends on the relative status of the identity it protects.
References
Acker, J. (1990). Hierarchies, jobs, bodies: A theory of gendered organizations. Gender & Society, 4 (2), 139-158. https://doi.org/10.1177/089124390004002002
Connell, R. W., & Messerschmidt, J. W. (2005). Hegemonic masculinity: Rethinking the concept. Gender & Society, 19 (6), 829-859. https://doi.org/10.1177/0891243205278639
Correll, S. J., Weisshaar, K. R., Wynn, A. T., & Wehner, J. D. (2020). Inside the black box of organizational life: The gendered language of performance assessment. American Sociological Review, 85 (6), 1022-1050. https://doi.org/10.1177/0003122420962080
Goffman, E. (1976). Gender advertisements. Macmillan Education UK.
Grip, L., & Jansson, U. (2022). ‘The right man in the right place’–the consequences of gender-coding of place and occupation in collaboration processes. European Journal of Women’s Studies, 29 (2), 250-265. https://doi.org/10.1177/13505068211042725
Khanna, K., & Meadow, T. (2023). The fragile male: An experimental study of transgender classification and the durability of gender categories. Gender & Society, 37 (4), 553-583. https://doi.org/10.1177/08912432231180152
Lunneblad, J., & Johansson, T. (2021). Violence and gender thresholds: A study of the gender coding of violent behavior in schools. Gender and Education, 33 (1), 1-16. https://doi.org/10.1080/09540253.2019.1583318
McDowell, L. (2004). Thinking through work: Gender, power, and space. In Reading economic geography (pp. 315-328). Blackwell. https://doi.org/10.1002/9780470755716.ch20
O’Brien, K., Petra, V., Lal, D., Kwai, K., McDonald, M., Wallace, J., Jeanmonod, C. & Jeanmonod, R. (2022). Gender coding in job advertisements for academic, non-academic, and leadership positions in emergency medicine. The American Journal of Emergency Medicine, 55, 6-10. https://doi.org/10.1016/j.ajem.2022.02.023
Oldford, E., & Fiset, J. (2021). Decoding bias: Gendered language in finance internship job postings. Journal of Behavioral and Experimental Finance, 31, 100544. https://doi.org/10.1016/j.jbef.2021.100544
West, C., & Zimmerman, D. H. (1987). Doing gender. Gender & Society, 1 (2), 125-151. https://doi.org/10.1177/0891243287001002002
