At first glance, the opinion page of any newspaper looks like a space for words and ideas, but a slower look at its headlines shows that it is also a space for faces, even if it contains not a single picture. One person enters this space carrying their name and what they said or did; another is given nothing in the headline but a general description, or a family tie to someone else. A headline, within its few words, is not a complete mirror of life or of the article it introduces, but it is a journalistic choice that carries meaning. So our question was not only how many women opinion writers wrote about, but how their presence is framed: by name or by description? As the subject of the sentence, or as the one acted upon? Defined independently, or attached to a man's name?
To get closer to an answer, we put every headline through a structured reading that identifies the people mentioned in it, the gender of each, how they are named, the role they are given in the sentence, and the tone in which they are discussed. One limit of what we will see should be noted from the outset: 72% of the headlines come from 2019 onwards, and the sources are concentrated in particular in Egypt, Qatar and the UAE. What we have, then, is a picture of the opinion pages of these newspapers during the years covered by the corpus, not a verdict on the Arab press as a whole; still, it is a picture broad enough to deserve a pause at each of its paradoxes, one after another.
The opinion page: a woman in every 28 headlines, a man in every 8
If we gathered these headlines on a single page and began to pick out those in which a man appears as the subject and those in which a woman does, a gap would open up before we read a single article. Of every hundred headlines, at least one woman appears in only 3.5%, while a man appears in 12.8%, or about one headline in every 8. Put another way, for every headline whose words are about a woman there are 3.6 about a man. Women's share even falls to 3.0% if we require the woman to be the focus of the headline rather than a passing mention. These calculations, here and throughout the investigation, do not count the names of women and men columnists merely for signing their articles.
Picture it: if a hypothetical opinion page were made on the scale of this corpus, where would the few lines featuring women fall among all the others?
By "a headline with a woman" we mean a headline that mentions a woman or women as its subject, not just the name of the woman who signed it. The headlines shown were chosen at random, after excluding any that might reveal the identity of private individuals.
The colon after the name… who is given the floor?
The colon that follows a name in a headline does more than punctuate: it opens the door to a voice being presented to the reader, "So-and-so writes:" or "So-and-so:", followed by the words that person is allowed to say or that are quoted from them. We looked for the owners of these voices, whether the article's authors or the people whose words the headlines quoted, and women's share was just 15%: 3,236 women against 18,480 men. This is not the share of women among all those who write for newspapers, but the share of women among those whose names appear in this particular way within the headline, a distinction whose precision should not be lost behind the appeal of the conclusion.
Imagine a hundred voices: if the colon were placed after a hundred names, how many of them would belong to women? And does the share look the same from one country to the next?
Here the geographic comparison shows a wide range, from 36% in Saudi Arabia down to 5% in Morocco. But the other surprise comes from what is said, not who says it: when the speaker is a woman, women are not necessarily her subject. Only 5.5% of women writers' headlines mention a woman, while 28% concern politics and war. A woman who is present through her voice is not confined to writing about women, and this is one of the paradoxes that could be lost if we only counted names.
The "speaker" category combines the columnist and the person quoted in the headline, and the corpus does not always distinguish between the two. Some countries are represented by only a few newspapers, so the share may reflect the policy of one or two papers rather than the situation of the whole country.
"Woman" tops the list of women… with no face and no name
When we ranked names and descriptions by how often they appear in headlines, the top of the women's list was a word that refers to no one in particular: "woman". It appeared in 752 headlines, 7.5 times as many as Umm Kulthum's name (100). The generic word "man", on the men's list, came no higher than 18th, as the list is led by the names of well-known presidents and leaders. This reveals a paradox worth considering: on one side the general description outranks the names, and on the other the names outrank the description, even though both are part of journalistic talk about human beings.
On the board of names: whom can we recognise when a man is mentioned? And who is the woman the headline does not even need to name?
Most frequent among women
Most frequent among men
Striped boards are general descriptions without a name; solid ones are people named by name. Lengths use the same scale in both columns.
Trump alone appears in 2,369 headlines, equal to 62% of all the headlines that named any woman by name, whoever she was (3,826 headlines). On the margins of the big names there is a small and deeply human paradox: "my mother" appears, in headlines whose writers celebrate or mourn their mothers, in 91 headlines, more than Fairuz (45). Between the name of fame and the name of kinship, it becomes clear that a woman's mere presence does not always tell us how she was imagined or introduced.
A few leaders crowd the top of the names
Although there are far more men mentioned by name, their names cluster more tightly at the top: the ten most frequent men account for 25.8% of all mentions of men by name, against 11.2% for the ten most frequent women. A direct comparison may be unfair here because of the difference in size, so we recalculated using an equal number of mentions, 4,015 on each side, and found 2,261 different women's names against about 2,009 men's names. In this comparison women's names are not less varied; they are less present overall, and less crowded around a small peak of celebrities.
Every mention of a woman by name
Every mention of a man by name
Each square = 1% of mentions by name; coloured squares are the share of the top ten names.
Different spellings of the same name, such as the Arabic variants of "Trump" and its Latin spelling, were merged for the names shown on the board only; the true rank of the word "man" may therefore be lower than the one shown.
When the man has a name… and the woman remains a description
A person has a name that sets them apart, but headlines do not grant names to everyone at the same rate. Of every hundred appearances of a man, 77% are by name, while only 40% of women's appearances are by name. The difference is not merely an overuse of titles: general words such as "woman", "women" and "girl" make up 28% of women's appearances, against only 4% for men. Here the name becomes an instrument of individuality, while the general description carries its subject into a wide box that could hold an endless number of women, without letting us recognise any one of them.
What cards do the faces carry? If a hundred men and a hundred women passed in front of us as the headline writers present them, how many could we call by name?
A hundred appearances of a woman
A hundred appearances of a man
The point is confirmed from another angle: in 28% of headlines that mention women, no particular woman appears at all, only women as a category whose role, rights or status is being discussed; the corresponding figure for men is 4%. It might be said that the explanation lies in different subjects, and that culture and society pages treat women in a way they do not treat male politicians and sportsmen. But when we gave men the same distribution of subjects as women, a man still appeared by name in about 64% of his appearances, against 40% for a woman. Differences between sections explain part of the gap, but they do not remove it.
This weighted comparison may soften the impact of the gap, but it does not reverse its direction. So it is not enough to say that women and men are spread across different sections; we must also look at the phrases chosen for each within the same section. Here the test is more precise, because it compares what readers are offered on similar subjects.
Even within the same subject, the names are not equal
When we look inside each section separately, from politics to sport, health and education, the rate at which women are named stays below men's in all thirteen sections (13 of 13). Only in the society and family section do the two come close, not because women's names become strikingly frequent, but because men are also named less. The gap, then, is not confined to where the headline talks about women; it extends to the form it chooses when it does.
The rate within each section is calculated from appearances by name for each gender, and only sections with at least a hundred appearances of each are shown.
A thread of kinship in the headline… whose is the name, and whose the tie?
Kinship is no flaw in an introduction; motherhood, fatherhood, marriage and siblinghood have their natural place in people's lives. But when a family tie becomes a way of defining one gender more than the other, the numbers show it: a woman is described as a mother, daughter, wife or sister in 9% of her appearances, 3 times the rate for men (3%). Of all the ties, marriage shows the sharpest imbalance: the word "wife" occurred 150 times and "husband" 41, a gap of 3.7 times. Here the person is not necessarily erased, but is often presented through a relationship with someone else.
The thread of identity: through what tie do the woman and the man enter the headline, and when does the tie become more important than the name?
Women by a family tie
Men by a family tie
This imbalance does not repeat in the same way across all family relationships: mentions of motherhood and fatherhood are close, 372 against 341. But the scales tip sharply when it comes to marriage: a man is rarely introduced as "So-and-so's husband", while well-known women are introduced in headlines as the wives of men who appear by name. These are examples of the actual wording we found, not headlines we made up to illustrate the difference:
When a woman's womanhood itself becomes the subject
A woman's name may appear in a headline without her being a woman having any direct bearing on the story, or her gender may be the very thing the whole sentence turns on. When we separated the two cases, womanhood turned out to be the focus in 52% of the headlines featuring a woman, whereas a man's manhood is a comparable focus in only 12% of men's headlines. In about half of these headlines, then, the meaning of the woman's presence depends on her being a woman, while in many of his headlines a man moves as a person whose deeds or words are considered without calling up his gender.
What makes the news? Is it what the woman did or said, or is the mere fact of her being a woman an essential part of what is being said about her?
A hundred headlines with a woman
A hundred headlines with a man
Coloured segments: headlines in which gender is the focus.
"Focus of the headline" is an estimate of how central gender is to the wording. This share compares headlines that mention a woman and no man with those that mention a man and no woman; it does not describe every kind of mixed headline.
In the grammar of the sentence… who acts and who is acted upon?
Words also have a motion as telling as their names and descriptions: the verb in a headline is assigned to one person and may fall on another. In 65% of men's appearances the man is the grammatical subject, deciding, attacking, winning or erring, while a woman is the subject in 48% of hers. Conversely, the action falls on a woman in 21% of her appearances, against 15% for a man. These are roles created by journalistic language and must not be equated with what real people actually did; but they tell us how the headline's wording distributed initiative and being acted upon.
If the headline were a stage: to whom is the action at the front of the scene given, and whom do we see receiving it?
Fifty women
Fifty men
When the two meet in one headline
In 1,312 headlines a man and a woman were mentioned together; the headlines named both in 560 cases. Where one was named and the other's name withheld, the man alone was named in 214 headlines and the woman alone in 47, a ratio of 4.6 to one. Looking at the direction of the action in those headlines, the man acts on the woman in 183 headlines and the woman on the man in 101. Bringing both genders into the same small space lets us see the gap in another way, but it does not remove the need to distinguish between the wording of the headline and the event it summarises.
The arithmetic paradox in shared headlines
Surprisingly, a man and a woman appearing together in a headline does not widen the naming gap as we might have expected. If we applied to the shared headline the naming rates seen when each appears alone (a man is named in 78% of his headlines, a woman in 39%), we would expect the man alone to be named about 5.5 times as often as the woman alone; but the observed ratio is 4.6. The gap narrows somewhat in shared headlines without disappearing. This is a descriptive result worth noting, not proof that the man's presence causes the woman to be named.
The number of headlines in which a man and a woman appear together is relatively small (1,312 headlines), so the differences here should be read as signals rather than very precise estimates.
The grammatical role is taken from the words of the headline alone: the "subject" is the one to whom the verb is assigned in the sentence, not a judgment on who took the initiative in the events themselves.
Warm words are not free of paradox
Who appears in the headline and in what capacity is not the only question; there is also the tone that accompanies the subject. We found that 44% of headlines mentioning a woman carry moderate or high emotion, against 33% of men's headlines and 21% of headlines that mention no people. Measuring positivity shows another gap: 26% for women against 13% for men. But it would be a mistake to confuse emotion or positivity with sensationalism: by this measure, headlines mentioning women are not the most sensational, at 22% against 24% for men's headlines.
But where does this warmth come from? Is it created by the culture and society subjects in which women appear, or by the way they are talked about within the same subject?
To answer, we went back to the sections, one by one, and the pattern did not disappear: women's headlines were more emotional in 12 of 12 subjects and more positive in 11. The type of subject alone is not enough to explain the gap; there is also a way of presenting women that leans towards tribute, mourning or pity. Conversely, questions appear in 12.2% of men's headlines, against 7.5% for women. This is no judgment that warmth is always good or that a question is always an accusation; it is a difference in wording that persists across subjects.
And when the evaluation is aimed at the person?
If we move from the tone of the whole headline to the way the person mentioned in it is described, we find women described more positively than men in 13 of 13 sections. The gap is widest on women's and gender issues and narrowest in history. The next panel lets you look at the share of positivity within each section, and then at the difference between the two shares; the difference between two groups is not the value of either one on its own.
The trace of the absent in the headlines' vocabulary
Even when we set the names aside and look at the words, the worlds of talk about men and about women were not alike. We compared the most distinctive words for each group after balancing the size of the two groups and excluding personal names and template words such as "writes", as identified by the analysis method; on the women's side, words such as "woman", "Egyptian (f.)", "mother", "rights", "women" and "for women" stood out, while on the men's side "man", "visit", "al-Assad", "Trump", "speech" and "son" stood out. The point is not that any of these words belongs to one gender, but that their relative presence is what distinguishes one side from the other in this measure.
Words have their fingerprints too: if the names step out of the scene, are the words alone enough to reveal how each side is presented?
Most distinctive of women
Most distinctive of men
Length = degree of distinctiveness, not the number of occurrences.
A map of the words that recur
Here we show the most frequent words in all headlines, in women's headlines and in men's headlines, after converting frequency to a rate per thousand headlines so that differences in size do not mislead us. Each word takes the colour of the section in which it appears most often; you can pause on any bar, click it, or search for a particular word. Length in the previous distinctiveness panel is not the number of occurrences; here we read a standardised frequency that can be compared. Words are shown in English translation.
Women between violence and empowerment… and appearance 8 times as often
In an earlier comparison of English-language headlines, a published analysis found that words of violence in headlines mentioning women outnumbered words of empowerment by about two to one. But the picture in these Arab opinion headlines is different: of every thousand headlines featuring a woman, 35 carry words from the field of violence, against 32 whose words relate to empowerment and achievement. Both categories are more frequent than in men's headlines. This is not a verdict on the full meaning of the headlines, but on the presence of words chosen in advance to measure these two fields.
The question of the words around women: where do the words of violence and of ability fall, and where does talk of the body and appearance show up?
The widest gap, however, is not between violence and empowerment but in words relating to appearance and the body: beauty, dress, the hijab, femininity and the like appear in 17 of every thousand headlines featuring a woman, and in only 2 of every thousand featuring a man, about 8 times as often. In one headline a female head of state is even described as "the most beautiful of the beautiful", an expression that puts appearance at the front of how a person whose presence also relates to public affairs is introduced. Here especially, we need to look at the word the headline used, rather than infer its writer's intention from it alone.
We used word lists we compiled for this measure; they detect the presence of a word, not the meaning it carried in context. "Hijab", for instance, may occur in a political debate. The comparison with The Pudding's English-language analysis published in 2022 is a comparison of the general direction only, not of the values of the shares, because the corpora and word lists differ.
In politics, one woman for every 11 men… while the family section comes close to parity
Moving from the individual headline to the newspaper's sections, the distance between the genders changes widely. In politics, women appear in only 1.7% of headlines, against 19% for men, roughly one woman for every eleven men in terms of their presence in these headlines. The society and family section is the only one in which the two shares come close to parity: 9.9% for women and 9.5% for men. In the "women's issues" section women appear in 74% of headlines, and 16% of all women's headlines go to this section. This is a picture of how appearances are distributed across subjects, not a measure of women's or men's participation in the actual events.
Across which sections are the names spread? And where does the distance between women's and men's presence shrink, and where does it widen?
There is another paradox that does not show up simply by searching for the word "woman" in a headline about women: we found 634 headlines about issues that directly affect women, from family law to harassment and motherhood, that do not mention a woman at all. This does not mean those articles lack women's voices or experiences; it means that the headline, which is the only object of this reading, addressed the issue without referring to a particular woman or to women as people present in it.
Identity cards: the politician on one side, the singer on the other
Professional titles and positions tell another story. When the job or professional capacity of the person in the headline is mentioned, "politician" tops the men's list with 5,371 occurrences, while "singer" tops the women's list with 243. Positions in government and politics account for 59% of the professions mentioned for men, against 25% for women. Art accounts for 17% of women's professional descriptions, against 3% for men, and family roles such as mother, wife and bride make up 11% on the women's side and no more than 1% on the men's. These are shares within the descriptions that occurred, not the distribution of professions among people outside the press.
If every name wore a card: what profession or capacity do headlines lean towards when introducing a woman, and which do they give a man?
Women's cards
Men's cards
The area of each card is proportional to how often the profession or description occurs in its column. The strip below each column divides a hundred mentioned professions or descriptions among their different fields.
Countries and years change… and no steady rise appears
Looking across countries and years does not give us a reassuring upward line. Women's presence in headlines ranges from 4.9% in Egypt to 1.8% in Qatar, and the differences between countries in the voices of women who write or are quoted appear wider than the differences in the women headlines are written about. In Egypt, women's presence rose from 1.8% in 2015 to 6.2% in 2022, then fell back to 4.5% in 2024. In Qatar it stayed close to 2%, while in the UAE it fell from 4.4% in 2005 to 1.3% in 2024. Because the newspapers represented and the size of the archive differ, not every change can be attributed to a real shift in society or in the press.
And what does time say? Do the years lead to a steady widening of women's presence, or does the picture rise and fall as the country and the archive change?
The archive of years
We compare years within the same country so as not to confuse a change in sources with a supposed trend over time. Each square represents a year, and its colour follows the share of headlines featuring a woman; the colour fades when the number of headlines is small.
What remains after reading these headlines?
- In every 28 headlines we find one featuring a woman, while we find one featuring a man in every 8.
- At the top of the women's list comes the word "woman", not the name of a particular woman; the general description is more frequent than the names of celebrities.
- A man is mentioned by name in 77% of his appearances and a woman in 40%, and kinship is used to introduce a woman 3 times as often as a man.
- The action is assigned to the man in 65% of his grammatical appearances and to the woman in 48%; these shares do not describe action in reality.
- Women's headlines are warmer and more emotional across subjects, without being more sensational, while words about appearance recur about 8 times as often.
- Women make up no more than 15% of those given a declared voice in the headline, and the spread of years does not trace a steady path of improvement.
It would not be fair to consider every one of these headlines biased, or to derive from the numbers a verdict on the intentions of their editors; the corpus does not reveal intentions, nor does it tell everything in the body of the articles. But it does let us see a linguistic habit that recurs until it becomes so familiar we no longer notice it: the man appears with his name and what he did, while the woman appears, more often, by her description, her kinship, or the appearance others see in her. If the headline is a narrow space in which the press chooses whom to name and how to present them, then those small choices, repeated across hundreds of thousands of headlines, deserve to be read slowly. Not to demand that every headline say everything, but to ask what it habitually says, and what it habitually leaves outside its words. It is also only fair to note that the idea for this investigation drew some of its features from the article "When Women Make Headlines" published on The Pudding, a work that opened an important door to examining how women appear in news headlines. The analysis of the Arab press presented here, with its corpus, measures, comparisons and findings, is independent research carried out entirely at Sia, based on our own data and research questions and using analytical tools we developed for the purpose. The inspiration lay in the general idea; the research and its findings are Sia's own original work.
Did you find this story useful?