Women and men in the United Nations: A corpus analysis of General Debate addresses

Abstract

The United Nations (UN) has demonstrated a commitment to women’s empowerment over seven decades of work. Yet gender inequality pervades in countries around the world and even within the UN. This corpus analysis investigated collocates of woman(’s), women(’s), man(’s) and men(’s) in 193 UN General Assembly General Debate addresses in fall 2015 to examine gender representation by international decision-makers. The analysis revealed that the plural women occurred more frequently than men and primarily in discussions on gender equality and violence. In contrast, the singular man was identified more frequently than woman, in generic use (e.g. mankind) and in references to eminent male leaders. Despite UN dedication to gender equality and linguistic guidelines promoting gender-inclusive language, these General Debate addresses do not conceptualise women and men equally, more frequently referring to women as a group in need of protection and support and to men as political figures and other distinguished individuals.

Keywords

Corpus-assisted discourse analysis gender international organisation discourse United Nations

Introduction

Since its inception, the United Nations (UN) has recognised the importance of advancing gender equality and empowering women. The Commission on the Status of Women, established in 1946 by the UN Economic and Social Council, has a mandate to monitor women’s rights and freedoms around the world and make suggestions for empowering women. The Universal Declaration of Human Rights, adopted by the UN General Assembly in 1948, acknowledges gender equality as a fundamental human right. More recently, the UN has adopted landmark agreements such as the Convention on the Elimination of All Forms of Discrimination Against Women (CEDAW); the Beijing Declaration and Platform for Action; Security Council Resolution 1325 on Women, Peace and Security; and Goal 5 of the Sustainable Development Goals on gender equality and the empowerment of women and girls. The UN Entity for Gender Equality and the Empowerment of Women (UN Women) monitors progress made toward the goals of these agreements and publishes reports every month on the status of women around the world.

Despite over 70 years of extensive UN work on women’s rights, gender inequality remains pervasive in every region of the world, developed and developing. Even within the UN, gender inequality prevails. Women make up 61.5% of the workforce at the lowest UN grade level (P-1) but only 27.3% at the highest (U-G, UN Women, 2016). Women are the majority only at the two lowest grade levels, P-1 and P-2, and their representation drops steadily with each increasing grade level. One of the most prestigious positions in the UN, the President of the General Assembly, has been held by a woman only four times out of 73. The highest-ranking position, UN Secretary General, has never been held by a woman. The discrepancy between (1) the emphasis placed by the UN on women’s equality and (2) the status of women working in the UN merits investigation, and could be better understood, then tackled, through an analysis of how women are represented in UN texts.

Such an investigation on gender and language in UN texts would contribute to a rich body of research. First, it would add to studies on language and gender, which have explored two principal domains: (1) differences in language use associated with gender and (2) ways in which gender bias is perpetuated through language. According to Lakoff (1973: 45) ‘The marginality and powerlessness of women is reflected in both the ways men and women are expected to speak and the ways in which women are spoken of’. Over the decades, the core tenants from Lakoff (1973, 1975) have been the object of discussion and re-evaluation (e.g. Crawford, 1995; Hall and Bucholtz, 1995; Lakoff, 2004; Tannen, 1994; Weatherall, 2002), and research continues to find evidence of differences in language use by women and men as well as gender bias reflected in statements made about women and men.

Within the domain of language and gender, numerous studies have investigated the realm of politics. These analyses have been varied in their methods and research questions. For example, Wodak (2003) used data from interviews to show that women working in the European Parliament identified themselves as different from their peers. Other studies have contrasted the communication styles of female and male politicians. Cameron and Shaw’s (2016) analysis of the political speech of three female party leaders in the United Kingdom revealed differences in both what female politicians say and how they say it. Still other investigations have looked at gendered language in political titles. In a study on gendered forms for the Italian title of mayor (e.g. sindaca, sindaco, candidate sindaco), Formato (2019) noted that each choice was ideological, reflecting either conformity to or subversion of the perceived norm. The overarching similarity in these studies is the finding that women in politics are typically viewed, by themselves and others, as outsiders in a realm historically reserved for men.

Another body of research on gender and language has examined the frequency, usage patterns, and collocates of gendered terms such as woman, man, girl, boy, spinster and bachelor (e.g. Baker, 2008, 2010; Caldas-Couthard and Moon, 2010; Holmes, 2001; Holmes and Sigley, 2002; Holmes et al., 2009; Macalister, 2011; Pearce, 2008; Taylor, 2013). These studies, on average, revealed a higher frequency of male-gendered terms (man, boys) compared to female (woman, girls) in texts from the United Kingdom, United States, Australia and New Zealand; however, diachronic investigations examining changes over time (e.g. Baker, 2010; Holmes et al., 2009) also found a decrease in the use of man/men and an increase in women and, to a lesser extent, woman. Holmes and Sigley (2002) noted that in their corpora of texts dating from the late 1980s to early 1990s, women is actually more frequent than men.

The importance of these findings lies primarily in the underlying reasons for frequency differences. The drop in use of man over the last few decades has been partly attributed to fewer instances of the generic use of man (e.g. mankind, Baker, 2010; Holmes, 2001) and to a shift toward more gender-neutral language for occupation titles (e.g. police officer rather than policeman, Baker, 2010; Holmes et al., 2009). A qualitative analysis in Holmes et al. (2009) revealed that the increase in women between 1961 and 1991 can be partially explained by the number of passages in their more recent texts about women’s equality and violence against women. The same was not true for the singular woman, which increased only slightly. Holmes and Sigley (2002) explained that women are discussed more frequently as a group than men. In contrast, individual women and subsequently the term woman are mentioned less frequently than individual men. They argued that true social progress in terms of gender equality would result in an equal number of texts about individual women and individual men. Holmes and Sigley further argued that a qualitative analysis is required to determine the substance of texts about men and women; they pointed out that references to individual women can be unflattering and sexist, so higher frequencies of female-gendered terms are not necessarily indicative of greater gender equality.

In addition to providing data on frequencies of gendered terms, many of these investigations have followed a Corpus-Assisted Discourse Studies (CADS) framework and undertaken collocation analyses. The objective of CADS is to explore how ‘language is used to (attempt to) influence the beliefs and behaviour of other people’ (Partington et al., 2013: 5) and to ‘uncover ideologies and evidence for disadvantage’ (Baker, 2006: 5). CADS investigate how politically or socially loaded terms (e.g. refugee, anti-Americanism) are conceptualised in different types of texts, often by examining collocates (i.e. words immediately preceding or following the target word), in order to shed light on how political, social or cultural beliefs are inculcated through language.

Corpus-assisted collocation analyses of gendered terms such as boys and girls have revealed a striking contrast between the conceptualisation of females, often depicted as vulnerable, and that of males, represented as powerful. In his investigation of the British National Corpus (BNC), Pearce (2008) found vulnerable, disadvantaged and submissive more frequently described women than men. Likewise, women were more likely to be the object of disempower and marginalise. In contrast, men were distinguished, eminent, great and influential. Though both women and men were often the grammatical object of physical violence, men were more likely to be the grammatical subject, the perpetrators of violent acts (assault, rape). In their study of British newspapers, Caldas-Couthard and Moon (2010) found that women were more likely to be depicted as powerless (vulnerable, traumatised) and to be sexualised (sexy, nude, topless). Baker (2010) examined gendered terms in four corpora comprising British press, general prose, academic writing and fiction, ranging from 1931 to 2006, and found that even in 2006, men were more frequently portrayed as venerable (powerful, celebrated, distinguished); women were more frequently described as desirable, voluptuous and sultry and not one instance of a great woman or an influential woman could be found (Baker, 2010: 138).

These gender representations have been identified in a wide variety of text types (e.g. written press, fiction, letters, academic prose, spoken conversation in both informal and formal settings, radio shows). Would portrayals of men and women be more balanced in UN speeches given the organisation’s dedication to empowering women? Might they also reveal some underlying inequality, reflecting the gender imbalance in UN employment, with women more commonly working at the lowest grades and men at the highest grades?

A range of studies have been carried out on the language of UN texts and have identified clear discourse, rhetorical, lexical and grammatical patterns in UN texts (e.g. Baturo et al., 2017; Brun-Mercer, 2018; D’Acquisto, 2017; Donahue and Prosser, 1997; Duchêne, 2008; McEntee-Atalianis, 2013). These patterns are typically followed by both UN officials (e.g. the Secretary General) and by national politicians such as heads of state delivering formal addresses before the UN.

The present corpus-assisted discourse study explores whether gendered terms in UN texts exhibit lexical patterns as well and to what degree these patterns adhere to or reject gender representations found in other text types. The investigation contrasts the conceptualisation of woman and the related words women, woman’s and women’s (hereafter WOMAN) with the conceptualisation of man, men, man’s and men’s (hereafter MAN), with a view to better understanding gender stereotypes and associations made in UN texts.

The texts selected were UN General Assembly General Debate (hereafter UNGA) addresses. The General Assembly is the primary policymaking body of the UN and each annual session opens with a General Debate in which presidents, prime ministers, foreign ministers and other high-ranking representatives make formal statements. The General Debate is a forum for outlining national opinion and action, recommending future steps at the international level and building and maintaining sound diplomatic relationships with fellow member states (Donahue and Prosser, 1997). The General Debate also offers the most representative sample of texts from speakers around the world because all 193 UN member states, along with three non-member observers (the European Union, the Holy See and Palestine), are invited to deliver a formal address and it is rare for a country to decline.

The corpus was restricted to 1 year to ensure sufficient data for quantitative analysis without being too large for detailed qualitative study. Both types of investigation were deemed important in order to find clear patterns as well as better understand the reasons for those patterns, including distinguishing between overt gender bias (e.g. men are strong, women are nurturing) and societal issues (e.g. women are more frequently victims of gender-based discrimination).

A final advantage of the corpus is the identification of each UNGA speaker, making it possible to investigate similarities and differences in how female and male speakers conceptualise WOMAN and MAN. Comparison by speaker gender is compelling because it adds to the literature on (1) self-identification and self-reference of female politicians and (2) conventionalised language in UN texts. Previous studies have shown differences in speech form and substance when comparing female and male politicians. However, given that UN texts have been shown to be highly formulaic, fewer differences may be found in UNGA texts.

Thus, this study aims to address the questions:

How are the terms WOMAN and MAN conceptualised in UNGA addresses?

What similarities and differences emerge in how female and male speakers conceptualise WOMAN and MAN?

Method

A corpus was compiled comprising all 193 UNGA addresses delivered by UNGA member states and observers during the fall 2015 General Debate of the 70th Session (28 September–3 October 2015). Addresses from 2015 were selected for the present study because women’s rights and advancement were of particular interest that year: The UN was celebrating the 20th anniversary of the Beijing Declaration and Platform for Action, which established measures for advancing the rights of women nationally, regionally and internationally. The year 2015 also marked the 15th anniversary of Security Council Resolution 1325 on Women, Peace and Security. In addition, on 25 September 2015, the UN adopted the 2030 Agenda for Sustainable Development, highlighting gender equality in Sustainable Development Goal 5: ‘achieve gender equality and empower all women and girls’ (United Nations General Assembly, 2015: 20).

Each UNGA General Debate address was downloaded from the United Nations General Assembly official records (http://www.un.org/en/ga/documents/pvsr.asp). Files were converted from pdf to plain text using Zilla, then cleaned to correct spelling errors resulting from reformatting (e.g. development becoming deveBopment) and to delete all unspoken words (e.g. cover page with name of country and speaker). Lexical and grammatical errors in the original pdf files were not corrected (e.g. crush rather than crash in Malaysian MH17 plane crush, Ukraine).

The final corpus comprises 383,500 words from the 193 addresses delivered in 2015. These texts include 190 of the 193 UN member states, as three UN countries did not speak in 2015 (Cameroon, Singapore and Uzbekistan), as well as texts from three observers (the European Union, the Holy See and Palestine). The corpus is composed of 99 texts delivered in English and 94 translated texts. Both originals and translations are considered official UN documents and placed in the UNGA official records. Translations, like documents that have not been translated, reflect the language that is being used in the UN and therefore, for the purposes of this study, the conceptualisation of WOMAN and MAN in UN texts.¹

To determine how WOMAN and MAN are conceptualised in UNGA addresses, collocates were identified using WordSmith Tools 7.0 (Scott, 2017). Collocates were operationalised as words occurring within a five-word span of the target term, the most common operationalisation (see Baker, 2006; Scott, 2017). Function words were eliminated and collocates were analysed semantically to find patterns in the types of associations UNGA countries made with WOMAN and MAN. The broader co-text surrounding these collocations was also examined qualitatively to elucidate further associations and explanations for those associations.

The second research goal is to identify any differences in the conceptualisation of WOMAN and MAN based on speaker gender. Speeches were categorised as either ‘male speaker’ (172 texts) or ‘female speaker’ (21 texts). Though the small dataset for female speakers makes generalisation impossible, the research question was retained as exploratory. Collocates were identified for the two groups and compared.

Analysis and discussion

Conceptualisation of WOMAN in the UNGA

WOMAN was identified in 118 texts (61% of the corpus) for a total of 340 tokens. The plural women(’s) occurred 320 times (94% of occurrences of WOMAN). To investigate the conceptualisation of WOMAN, the collocates of WOMAN were identified across all UNGA texts. Table 1 shows the top 20 content word collocates of WOMAN, with information on token frequency, distribution (texts), and strength of association (Mutual Information, MI). Higher MI scores indicate stronger association, with MI = 3.0 conventionally considered the cut-off for significant collocates (Hunston, 2002: 71).

Table 1.

Top 20 content-word collocates of WOMAN for all UNGA.

Collocate	Tokens	Texts	MI
Men	55	43	10.03
Children	46	39	8.53
Empowerment	34	25	12.86
Girls	33	23	9.73
Rights	32	24	5.80
Peace	30	20	4.87
Security	28	17	4.77
Equality	24	20	10.81
Gender	18	15	10.85
Youth	14	11	8.24
Violence	14	10	6.66
People	13	12	3.98
Resolution	13	12	6.35
World	13	13	3.55
UN	12	10	4.16
Participation	12	11	10.71
Vulnerable	10	9	6.96
Role	8	8	5.58
Equal	8	8	7.33
Elderly	8	8	9.51

The top 20 collocates suggest a relative homogeneity in how WOMAN is conceptualised in UNGA addresses: a dozen collocates were identified in 10% or more of the texts using WOMAN, a high rate of dispersion. Men, the most frequent collocate, was identified in 43 texts (36% of the 118 texts using WOMAN) for a total of 55 co-occurrences (16% of all tokens of WOMAN). The dispersion of the collocate men across the corpus and its overall frequency reveal a strong tendency for women to be associated with men.

The top 20 collocates of WOMAN can be categorised into five semantic domains: (1) equality (empowerment, rights, equality, gender (equality), participation, role, equal), (2) vulnerability (violence, vulnerable, gender (-based violence)), (3) UN Security Council Resolution 1325 on Women, Peace and Security (peace, security, resolution, UN), (4) UN purview (UN, world) and (5) people (men, children, girls, youth, people, elderly). These semantic domains are discussed in turn with textual examples.

Women and equality

Of the top 20 collocates, seven suggest an association between WOMAN and equality. Numerous speakers mention, for instance, the empowerment of women, women and equal rights, and the participation or role of women in government and peacekeeping. Extracts 1 and 2 illustrate the relationship between WOMAN and equality. In all text excerpts, the target word (WOMAN or MAN) is italicised and collocates are in bold.

Extract 1. The Netherlands on WOMAN and equality

There is still not enough focus on the positive role that women can play. Even in the harshest and most hopeless circumstances, women often have the strength to go on. To find practical ways to improve the lives of their families and communities. Promoting equal rights and opportunities for women is actually sustainable development in action.

Extract 1 from the Netherlands emphasises the importance of equality and the participation of women, though the only details regarding the exact role they play are in practical ways to improve the lives of their families and communities. No mention is made of women’s participation in the highest echelons of national or international decision-making. In contrast, Extract 2 from Mongolia refers specifically to women holding more positions of power.

Extract 2. Mongolia on WOMAN and equality

We believe in gender equality and women’s empowerment. Mongolia is a strong supporter of the UN in this cause. For societies to advance, we need more women in public service at all levels - local and global. If women hold more positions of power, we will have less suffering and conflict, and more harmony and civic engagement.

In addition to Mongolia, several other countries (e.g. Malawi, Suriname, Vanuatu) call for greater representation of women in both national and international political arenas. Still other countries distinguish themselves by detailing national progress made in women’s political participation (e.g. Women now claim 11 percent of judgeships, Afghanistan).

Women and vulnerability

Calls for the empowerment of women relate not only to strengthening women’s political and social roles but also to improving their physical safety. In dozens of texts, women are represented as victims of violence, with vulnerability-related collocates such as violence and vulnerable (see Extract 3).

Extract 3. Ireland on WOMAN and vulnerability

I wish to emphasise Ireland’s grave concern for vulnerable women and girls in many parts of the world and our strong commitment to gender equality. On his visit to Ireland in May, Secretary General Ban spoke passionately of his personal commitment to gender equality and the UN’s efforts to combat the emergence of appalling forms of sexual violence and enslavement. Mindful of the 20th anniversary of the Beijing Conference, we must keep women’s rights high on our agenda.

Extract 3 from Ireland deplores the physical vulnerability of women, associating the commitment to gender equality with combating sexual violence.

Women and Security Council Resolution 1325

Many references to violence against women are made in the context of Security Council Resolution 1325 on Women, Peace and Security, which acknowledges the particular vulnerability of women in situations of conflict (see Extract 4).

Extract 4. Republic of Korea on WOMAN and Resolution 1325

This year marks the 15th anniversary of the adoption of Security Council resolution 1325 on women, peace and security, and the international community should do justice to the occasion by paying greater attention to sexual violence against women in conflict situations.

Security Council Resolution 1325 also calls for greater participation of women in conflict resolution and several texts emphasise the importance of including women in peacekeeping operations, as shown in Extract 5 from Slovakia.

Extract 5. Slovakia on WOMAN and Resolution 1325

It has been 15 years since the adoption of the landmark Security Council Resolution 1325 on Women, Peace and Security. . . . The presence of women negotiators in high-profile United Nations peace and mediation processes, role of women deployed by the United Nations to major conflict zones and highstake inter-state negotiations, as well as in peacekeeping and peace-building in general, is irreplaceable and should continue to grow.

Extracts 4 and 5 illustrate how the phrase women, peace and security refers frequently to the title of Security Council Resolution 1325. Of the 30 co-occurrences of peace and the 28 co-occurrences of security with WOMAN, 17 are in the phrase Women, Peace and Security. In 11 of the 12 co-occurrences of resolution and WOMAN, the reference was to Resolution 1325. The other resolution also related to conflict (United Nations General Assembly Resolution on “Women, Disarmament, Non proliferation and Arms Control”). In three instances, UN collocated with WOMAN in references to UN Resolution 1325.

Women and UN purview

UN collocated with WOMAN in the name of the organisation UN Women as well, and is therefore linked to UN purview. World was also categorised in this semantic domain, as it referred to the mission of the UN (e.g. we cannot transform our world unless the place of women within it is transformed, Australia), the scope of the UN (e.g. most of the world ’s poorest people are women, Ghana) and UN conferences (e.g. World Conference on Women, Cote d’Ivoire).

Women and people

The final semantic domain of WOMAN collocates comprises ‘people’ terms (men, children, girls, youth, people, elderly). These ‘people’ collocates are often enumerated in lists of vulnerable or disadvantaged populations, as shown in Extract 6.

Extract 6. San Marino on WOMAN and people

The Republic of San Marino has always paid special attention to the most vulnerable groups, such as women, children, the elderly and the disabled. Today, women are still the victims of discrimination and violence in many parts of the world, including in the most developed countries. Trafficking in women is far from being solved. Women living in conflict and post-conflict situations are often subject to sexual violence, torture and summary executions.

Some of the ‘people’ related collocates reveal a focus on both women’s vulnerability and their role in the family (see Extract 7).

Extract 7. Somalia on WOMAN and the family

Our women - mothers, daughters, sisters - also deserve limitless praise for enduring the worst of our brutal conflict. The survival instinct of the Somali people is most evident in women: raising families whilst their homes are being bombed, walking miles without water in search of safety and burying their children whilst continuing to work to feed the rest of their family. We know what war can do. And that is why I will ensure – as a father and a husband – that we will defeat terrorism. It is our duty, as a Government.

In Extract 7, the government, represented by a father and husband, has a duty to defeat terrorism to alleviate the plight of women during conflict. A qualitative examination identified nine texts (Bosnia and Herzegovina, Burundi, Ghana, Italy, Japan, the Netherlands, Poland, the Seychelles and Somalia, representing 5% of the UNGA) that refer specifically to women as mothers.

Most co-occurrences of WOMAN with ‘people’ terms relate to vulnerability. However, girls and men also co-occur with WOMAN in discussions on equality. Extract 8 refers to both violence against and the rights of women and girls.

Extract 8. Estonia on WOMAN and girls

This year marks the 20th anniversary of the Beijing Declaration and the Platform of Action that have influenced how we address equality and women’s rights today. We need to continuously stand for the rights of women and girls and to strive for the elimination of gender based violence. Women’s rights, and the empowerment of women and girls should be promoted at all levels.

Likewise, WOMAN and men co-occur in both passages on vulnerability to violence (e.g. children , women and men arriving in their thousands on our doorsteps in the Balkans as they seek safety and protection in Europe, Albania) and passages on equal rights (e.g. the vast inequality between men and women that so many traditions have inculcated, Ghana). Additionally, several co-occurrences of men and WOMAN are in reference to peacekeepers (e.g. men and women in uniform, Sierra Leone), or people more generally (e.g. good men and women, Jordan).

The order of WOMAN and MAN in these texts is noteworthy. Motschenbacher (2013) found men and man more frequently preceded women and woman in binomials in the written BNC. Though this preference for putting MAN first could be attributed to phonology or orthography, with initial placement favored for the shortest term in the binomial, arguably, word order in binomials can represent a social or cultural order (see Motchenbacher for a discussion). In the present study of UNGA texts, MAN is much more frequently first, appearing 45 times before WOMAN and only 13 times after WOMAN. Of particular interest is the context when WOMAN precedes MAN: in five passages on humanity (e.g. the world is made up of women and men with dreams and hopes, Belgium) and in eight passages on vulnerability (e.g. gender-based violence perpetrated against girls and boys, women and men , Costa Rica). In all nine passages on equality and in all nine passages on people serving in peacekeeping operations, MAN came first (e.g. Finland has contributed 50,000 men and women to UN peacekeeping operations, Finland).

Phonology is unlikely to account for this word order because it does not affect the order of other binomials such as women and girls. Of the 34 instances where girls collocates with WOMAN, girls, the one-syllable word, precedes women only twice. The initial position of men in men and women could reflect an underlying social order and also implicitly convey the expectation of finding men first and foremost in some public arenas such as peacekeeping.

Finally, under the ‘people’ domain, WOMAN was occasionally associated with specific individuals in passages that applauded their work. Because these passages were relatively rare (in only three texts) and did not make use of the same terms (e.g. only one occurrence of exemplary women), no collocates helped to identify this representation of WOMAN. Instead, the few references to great women were found in a qualitative examination of the data. Twice WOMAN is used in passages about specific female leaders. In the text from Liberia: At seventy only three women have served as President of the General Assembly, one of them being our compatriot Mrs. Angie Brooks Randolph and in the text from Ecuador: Our region is home to such exemplary women as Michelle Bachelet, Cristina Fernández, Dilma Rousseff, María Emma Mejía Vélez, Alicia Bárcena Ibarra and others.

The scarcity of references to great individual women in the UNGA explains the infrequent tokens of WOMAN in the singular (6% of all WOMAN tokens). The comparatively high frequency of the plural women(’s) reflects the UNGA propensity to discuss women as a group. Women are generally not portrayed as powerful individuals and leaders but are often designated as a group in need of protection and empowerment.

Conceptualisation of MAN in the UNGA

MAN was identified in 70 texts (36% of the corpus) for a total of 111 tokens, less than a third of the number of occurrences for WOMAN. Interestingly, 72 of these tokens (65%) were the plural men(’s) while 39 were the singular man(’s). Thus, men is less frequent than women (72 tokens compared to 320) but man is nearly twice as frequent as woman (39 compared to 20). Table 2 lists the top 20 content word collocates of MAN.

Table 2.

Top 20 content-word collocates of MAN for all UNGA.

Collocate	Tokens	Texts	MI
Women	55	43	10.06
Children	13	11	8.68
Made	10	10	7.39
Rights	7	6	5.58
Equal	6	6	9.11
Millions	6	4	8.08
Disasters	5	5	9.83
Nations	4	4	3.85
Human	4	4	4.49
Natural	4	4	7.49
Girls	4	4	8.61
Boys	4	4	10.81
Peace	4	3	3.99
Thousands	3	3	6.92
Countries	3	3	3.90
People	3	3	4.84
Earth	3	2	9.59
Beings	3	2	8.25
Young	3	3	8.79
Great	3	3	7.05

Of the top 20 collocates of MAN, 18 can be categorised into five semantic domains: (1) equality (rights, equal), (2) UN purview (millions, nations, peace, thousands, countries, earth), (3) humanity (man-made), (4) people (women, children, human beings, girls, boys, people, young) and (5) eminent individuals (great). The remaining collocates, natural and disasters were typically used as the lexical phrase natural disasters in contrast to man-made. Though identified in passages where MAN referred to all humanity, they were left unclassified because they did not explicitly relate to humanity. Overall, the semantic domains associated with MAN are similar to those with WOMAN. This is not surprising because, of the 111 occurrences of MAN, 55 co-occur with women and many of the other occurrences of MAN are in passages about women, even though WOMAN does not figure within a five-word span of MAN.

Men and equality

One similarity in semantic domains is the association of both WOMAN and MAN with gender equality (rights, equal). However, unlike for WOMAN, no references are made to the empowerment of men or for the need to promote the role or participation of men in political processes. Rather, men are mentioned as the other side of the gender equality coin. All six co-occurrences with equal relate to equality between men and women, with five in the phrase equal rights of men and women. The focus remains on the need to help women, not men, achieve equality, as shown in Extract 9.

Extract 9. Trinidad and Tobago on MAN and equality

Regrettably, in some parts of the world, women and children continue to be denied fundamental human rights and freedoms, receive less pay for equal work as men, and are systematically prevented from obtaining an education, all of which hinder their ability to participate in the sustainable development of their countries.

Men and UN purview

A number of collocates relate to the UN mission to represent and support all the countries of the world (e.g. It is a great honour for me to take the floor before the Assembly of peoples to celebrate, with men and women of all countries , races, languages and cultures, the seventieth anniversary of the founding of the United Nations, Djibouti). Terms such as millions and thousands often refer to the people served by the UN or serving in the UN (e.g. I would like to take this opportunity to recognize the efforts of the thousands of men and women of this Organization, Panama). In passages with UN purview collocates, MAN either co-occurs with WOMAN or represents all humanity (e.g. since man first appeared on Earth , Comoros) so the UN purview is not associated with males per se, but rather with all humanity.

Men and humanity

MAN also represents all people in expressions such as mortal men (Saint Vincent and the Grenadines), the co-existence of man and nature (Republic of Korea) and in the best interest of men and their communities (Serbia). Nine of the occurrences where MAN signifies humanity are in the phrase man-made (e.g. man- made disasters , Gambia; man- made crisis, Australia; and man- made catastrophe, Mali). In contrast, WOMAN is never used to suggest all of humanity.

It should, however, be noted that these expressions are eclipsed by the more frequent gender-inclusive terms in the UNGA. Humankind, for instance, occurs in 63 texts for a total of 125 tokens. The UN has published guidelines for gender-inclusive language, noting that ‘gender-inclusive language is a powerful way to promote gender equality and eradicate gender bias’ (https://www.un.org/en/gender-inclusive-language/index.shtml). These guidelines specifically call for the replacement of mankind with humankind, humanity, or human race and man-made with artificial or human-caused (see https://www.un.org/en/gender-inclusive-language/guidelines.shtml).

While the tendency is for gender-inclusive language (e.g. humanity), MAN was nonetheless used to represent all people in the UNGA by 12 speakers delivering texts in English and by the translators of 13 texts, including UN translators working out of Arabic, Chinese, French and Spanish. This use of MAN was identified in other text types (Baker, 2010; Holmes, 2001), but it is more surprising in UN texts, where diplomacy implicitly and guidelines explicitly call for careful attention to inclusive language.

Men and people

The largest semantic domain results from MAN co-occurring with other people-related terms. Like with many of the collocates in other semantic domains, women, children, human beings, people and young co-occur with MAN when men are not being singled out as males. Typically, these ‘people’ terms were identified in passages where men are acknowledged as victims of violence, but systematically alongside women and, often, children. Thus, in some texts women and children (but not men) are presented as vulnerable to war and conflict; in other texts men, too, are included among the victims. In no text, however, are men and children described as vulnerable without the presence of women as well.

In a few passages with ‘people’ collocates, speakers advocating an end to gender violence present females as victims and males as aggressors. Extract 10, for instance, mentions two projects intended to include men and boys in putting a stop to violence against women and girls: the ‘he-for-she campaign’ and the ‘boys-to-men’ project.

Extract 10. Zambia on MAN and people

Your Excellencies, one of the important programs in which I have been engaged as a designated champion of the “he-for-she campaign” seeks to address issues of gender based violence. The “he-for-she” project was initiated by UN women in September last year to focus on addressing the incidences of gender-based violence. In a few days’ time my government will again be launching the US aid supported “boys-to-men” project targeting the 15–21 year old age group and aimed at encouraging the development of a non-violent generation.

In this passage on ending violence against women and girls, boys and men are perceived as being the potential agents of violence while women and girls are potential victims.

Eminent individuals

Only one collocate, great, was identified for the category ‘eminent individuals’, but a qualitative examination of the data revealed the importance and distinct nature of this use of MAN. In addition to great describing remarkable men (e.g. a great man Nelson Mandela, Belarus), nine other references to specific notable individuals were identified without the collocate great (e.g. I remember the words of the young Muslim man Lassana Bathily who saved several innocent lives during the hostage crisis earlier this year in Paris, Bulgaria). These men are characterised as heroes or exceptional leaders. In the text from India, Gandhi is an extraordinary man with a powerful message. In the text from Mali, Pope Francis is a man of peace and a man of heart. In all, 11% of MAN tokens refer to specific eminent individuals, compared to less than 2% of WOMAN tokens. This preference for speaking of men as individuals helps account for the fact that the singular man(’s) represents 35% of MAN tokens while the singular woman(’s) represents only 6% of WOMAN tokens.

Summary of UNGA findings

Overall, MAN is frequently associated with gender equality and vulnerability to violence, but in every association with vulnerability and in nearly all the associations with equality, MAN co-occurs with WOMAN. When WOMAN is not a collocate, WOMAN (or a related word such as female) is present in the larger co-text. In contrast, WOMAN is often associated with gender equality and vulnerability without MAN in the co-text. Thus, gender equality and vulnerability are presented as issues related to (1) both women and men or (2) just women, but never related to just men. MAN can represent all of humanity (32% of all tokens of MAN); WOMAN never does. Of all tokens of MAN, 11% refer to great leaders in contrast to 2% for WOMAN. These findings are in keeping with previous studies (e.g. Baker, 2010; Holmes and Sigley, 2002; Holmes et al., 2009), but are somewhat more surprising in UN texts in which an emphasis on gender equality is presumed.

Representation by speaker gender

The question remains as to whether female speakers in the UNGA speak more frequently about women or conceptualise WOMAN differently. Table 3 summarises dispersion and frequency findings for the entire UNGA and for the two speaker groups. Overall, no significant difference was found in the number of texts with references to WOMAN or MAN when comparing female and male speakers. The difference in the number of tokens of MAN was more notable, but still not significant. Only the frequency of WOMAN was significantly greater for female speakers, but 49 of the 51 tokens were in the plural women(’s), indicating that while female speakers used WOMAN more frequently than their male counterparts, they spoke of women as a group rather than individual females.

Table 3.

Frequency of WOMAN and MAN in UNGA and by speaker gender.

Group	Female speakers	Male speakers	All UNGA	χ²	p
Dispersion
Texts using WOMAN	16	102	118	1.59	0.207
Texts using MAN	11	59	70	1.92	0.166
Total number of texts	21	172	193
Frequency
Tokens of WOMAN	51	289	340	7.16	0.007
Tokens of MAN	18	93	111	3.40	0.065
Total word count	39,965	343,535	383,500

χ² and p values given with Yates’ correction factor for one degree of freedom.

Conceptualisation of WOMAN by speaker gender

Though female speakers tend to talk about women and, to a lesser extent, men more frequently, the substance of the passages on women and men is not, on average, notably distinct. For WOMAN, 8 of the top 10 collocates are shared by female and male speaker groups (see Table 4). Proportionately more women speak about Resolution 1325 on women, peace and security, but no meaningful differences were detected either in the most frequent collocates of the two groups or in a qualitative examination of the texts.

Table 4.

Top 10 content-word collocates of WOMAN by speaker gender.

Female speakers	MI (female)	Male speakers	MI (male)
Peace*	6.00	Children*	8.66
Security*	6.00	Men*	9.82
Men*	9.58	Rights*	5.94
Girls*	9.22	Empowerment*	13.02
Empowerment*	11.86	Girls*	9.74
Violence	7.33	Equality*	10.91
Resolution	7.55	Security*	4.37
Children*	7.48	Peace*	4.43
Equality*	7.29	Gender	11.08
Rights*	4.81	Youth	8.31

Denotes collocate shared by both groups.

Some of the female speakers spoke passionately about women’s rights but the same was true for some of the male speakers. Extracts 11 and 12 are two of the most fervent passages about empowering women, both advocating a female Secretary General. Extract 11 from Liberia was delivered by a woman and Extract 12 from Costa Rica by a man.

Extract 11. Liberia, fervent female speaker on gender equality

This year we have marked the twentieth anniversary of the historic Beijing Conference and its Plan of Action for Women’s empowerment and gender equality. We are also marking the 15th anniversary of resolution 1325 on women’s participation in peace processes. And while improvement has been made in the status of women, much more is still to be done. We must therefore all “step-up” and take further measures beyond moralizing gender equality to tangible actions. . . . Only a few women serve as special representative of the Secretary General and not a single woman has ever served as Secretary General.

Extract 12. Costa Rica, fervent male speaker on gender equality

Costa Rica is proud to have included, in the resolution that was approved just a few weeks ago, an invitation to member States to present women candidates for the position of Secretary General. The time has come for the General Secretariat to be occupied by a woman. Let us acknowledge the great capacity, competence and commitment of women in all areas, and let us send an unequivocal political signal that, as we commemorate the twentieth anniversary of the Beijing Declaration and Platform for Action, this organization stands by what it preaches in terms of equality and empowerment of women and girls.

These two texts stand out for the vehemence of their calls for better representation of women in the UN. Only 4% of UNGA speakers (Bulgaria, Costa Rica, Ecuador, Iceland, Ireland, Liberia and the Philippines) mention the scarcity of women in the highest UN positions. More UNGA speakers refer to the role of women as mothers than to women as competent professionals capable of heading the UN as Secretary General or the General Assembly as President.

Furthermore, both female and male speakers mentioned women’s issues in passing without expressing particularly grave concern (see Extract 13 from a female speaker and Extract 14 from a male speaker).

Extract 13. Chile, female speaker with cursory reference to gender equality

Equality within and among countries; between men and women; growth, industrialization and inclusive and sustainable consumption, protection of the environment in its various facets, social peace–to name but a few of the Agenda’s goals–these are the face and the foundation of a mode of coexistence based on very profound values.

Extract 14. Algeria, male speaker with cursory reference to gender equality

Algeria has implemented an ambitious development policy and development plans whose purpose is to relaunch the economic growth and to provide for the social needs of all citizens, while continuing to strengthen the rule of law, independence of the judiciary, freedom of expression and equal opportunity for men and women.

Overall, no substantive differences were identified in the ways female speakers and male speakers conceptualised WOMAN. Female speakers did not refer to great female individuals more than male speakers and, similar to their male counterparts, focused primarily on women as a group in need of protection and support.

Conceptualisation of MAN by speaker gender

Likewise, female speakers used MAN in much the same way as male speakers. As shown in Table 5, five of the seven content-word collocates identified for female speakers were also identified for male speakers. (Because of the small number of texts by female speakers, only seven content-word collocates were identified.)

Table 5.

Top content-word collocates of MAN by speaker gender.

Female speakers	MI (female)	Male speakers	MI (Male)
Women*	9.73	Women*	10.06
Made*	8.75	Children*	8.71
Nations	5.98	Made*	7.15
Natural	9.40	Millions	8.27
Equal*	10.25	Equal*	8.52
Rights*	6.19	Rights*	5.38
Children*	8.44	Disasters	9.41
		Human	4.77
		Boys	10.74
		Young	9.05

Denotes collocate shared by both groups. Only seven collocates were identified for female speakers.

In 10 of the 18 occurrences of MAN in texts delivered by female speakers, it co-occurs with WOMAN. The co-occurrence of MAN and WOMAN relates to gender equality in three instances, to violence and vulnerability in four instances and to people in general in three instances (e.g. the men and women who may have planted its seed, India). MAN is also used to represent all humanity by female speakers five times (e.g. the aspiration of all men for peace and justice, Bangladesh) and MAN is used by women in reference to specific great men three times (e.g. a man whom I greatly admire, Liechtenstein, on Benjamin Ferencz). The use of MAN among female speakers is therefore similar to the representation of MAN for the UNGA as a whole.

Given the small number of texts delivered by female speakers, generalisation is impossible. However, the overall tendencies show that no meaningful differences distinguish how female and male speakers conceptualise WOMAN and MAN.

Conclusion

The aim of this study was to better understand how UN member states conceptualise WOMAN and MAN in UNGA texts. Associations with these terms were relatively consistent across texts, even when comparing female and male speakers. Findings also revealed some similarities between these 2015 General Debate addresses and corpora used in other gender-term studies: because of the focus on women as a group, particularly in discussions on equality and vulnerability, the plural women occurred more frequently than men. In contrast, the singular man occurred more frequently than woman because there were more references to exceptional men than exceptional women and because of the high frequency of man signifying humanity.

The finding that these UN addresses exhibit strong patterns in the associations made with WOMAN and MAN supports previous studies showing UN language to be highly conventionalised. Particularly striking is the lack of meaningful differences based on gender, given that previous research on gender, language and politics has shown notable differences in self-identification, self-representation and language use among female politicians. Considering the small dataset for female speakers in the present study, further investigation is needed, but it appears that language use in the General Debate is associated more closely with UN institutionalised expectations rather than gender.

The present study also contributes to research on the ways in which language can be used to influence beliefs about gender. These UNGA texts avoid some examples of gender bias found in other registers (e.g. women portrayed as sexy). However, as in studies of other registers, women are frequently associated with gender-based inequality and violence. This is likely due not to gender bias on the part of UNGA speakers but rather to the social issues of concern to the UN. Though a distinction must be made between (1) institutional bias and (2) social issues being addressed by the institution, future studies might explore the ways in which emphasizing certain social issues while downplaying others can contribute to gender bias (e.g. always referring to women as victims can solidify the image of women as victims rather than empowered leaders).

Two less overt types of gender bias found in this study are more concerning. First, the departure from gender-inclusive language by both male and female speakers, in translated and English-original texts, reveals that speakers and even official UN translators are not always following institutional language guidelines. As noted in Lakoff (1973: 76), ‘social change creates language change, not the reverse’. Thus, while UN policy on gender-inclusive language is laudable, these guidelines will not necessarily change perceptions of gender; moreover, language change might not be evident or widespread in practice until attitudes have changed sufficiently for speakers to be receptive to the change.

Even more disquieting is the relatively infrequent representation of women as heroes and leaders. While differences in representation of women and men may be somewhat less surprising in corpora of fiction or even newspapers, they are (or should be) incongruous in texts delivered at the UN, with its emphasis on gender equality. Frequent references to the achievements of individual male, but not female, leaders perpetuate underlying beliefs about who is suitable for political leadership. It is hoped that General Debate speakers will begin referring more to eminent female leaders such as Gro Harlem Brundtland, Jeanne Martin Cissé, Vijaya Lakshmi Pandit and Mary Robinson, reflecting a UN in which women, too, are portrayed as competent professionals qualified to work in the highest ranks of the UN, including the most prominent position as Secretary General.

In addition to contributions to the literature on UN discourse and on gender and language, this study underscores some important considerations for CADS and other corpus-informed work. Quantitative findings detect patterns and compelling avenues for further, qualitative investigation. However, without a careful manual examination of the texts, the reasons for these patterns are not always clear. Qualitative analysis can also uncover patterns that move beyond lexemes, patterns such as proper nouns being associated more frequently with MAN in the present study. It is hoped that gender in UN texts will continue to be studied through both quantitative and qualitative methods to better understand gender representation in the UN and potentially help guide efforts to promote gender equality in the world as well as within the UN itself.

Footnotes

Declaration of conflicting interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

Notes

Author biography

Nicole Brun-Mercer holds a PhD in Applied Linguistics from Northern Arizona University, Flagstaff. She currently teaches at the University of California, Berkeley. Her research interests include corpus linguistics, lexicogrammar, reading, and composition. Her publications have appeared in a variety of edited collections and journals, including TESOL Quarterly and The CATESOL Journal.

References

Baker

(2006) Using Corpora in Discourse Analysis. London: Continuum.

Baker

(2008) “Eligible” bachelors and “frustrated” spinsters: Corpus linguistics, gender and language. In: Sunderland

Harrington

Stantson

(eds.) Gender and Language Research Methodologies. London: Palgrave, pp.73–84.

Baker

(2010) Will Ms ever be as frequent as Mr? A corpus-based comparison of gendered terms across four diachronic corpora of British English. Gender and Language 4(1): 125–149.

Baturo

Dasandi

Mikhaylov

(2017) Understanding state preferences with text as data: Introducing the UN General Debate corpus. Research & Politics 4(2): 1–9.

Brun-Mercer

(2018) Nations united through discourse: A corpus analysis of UN General Assembly addresses. PhD Dissertation, Northern Arizona University.

Caldas-Couthard

Moon

(2010) ‘Curvy, hunky, kinky’: Using corpora as tools for critical analysis. Discourse & Society 21(2): 99–133.

Cameron

Shaw

(2016) Gender, Power and Political Speech: Women and Language in the 2015 UK General Election. Basingstoke: Palgrave Macmillan.

Crawford

(1995) Talking Difference: On Gender and Language. London: Sage.

D’Acquisto

(2017) A Linguistic Analysis of Diplomatic Discourse: UN Resolutions on the Question of Palestine. Cambridge: Cambridge Scholars Publishing.

10.

Donahue

Prosser

(1997) Diplomatic Discourse: International Conflict at the United Nations – Addresses and Analysis. Greenwich: Ablex Publishing Corporation.

11.

Duchêne

(2008) Ideologies Across Nations: The Construction of Linguistic Minorities at the United Nations. Berlin: Mouton de Gruyter.

12.

Formato

(2019) Women in the public sphere: Gendered language. In: Formato

(ed.) Gender, Discourse and Ideology in Italian. Basingstoke: Palgrave Macmillan, pp.135–198.

13.

Hall

Bucholtz

(1995) Gender Articulated: Language and the Socially Constructed Self. New York, NY: Routledge.

14.

Holmes

(2001) A corpus-based view of gender in New Zealand English. In: Hellinger

Bussmann

(eds.) Gender Across Languages: The Linguistic Representation of Women and Men, vol. 1. Amsterdam: John Benjamins, pp.115–136.

15.

Holmes

Sigley

(2002) What’s a word like girl doing in a place like this? Occupational labels, sexist usages and corpus research. In: Peters

Collins

Smith

(eds.) New Frontiers of Corpus Linguistics. Amsterdam: Rodopi, pp.247–263.

16.

Holmes

Sigley

Terraschke

(2009) From chairman to chairwoman to chairperson: Exploring the move from sexist usages to gender neutrality. In: Peters

Collins

Smith

(eds.) Comparative Studies in Australian and New Zealand English. Amsterdam: John Benjamins, pp.181–202.

17.

Hunston

(2002) Corpora in Applied Linguistics. Cambridge: Cambridge University Press.

18.

Lakoff

(1973) Language and woman’s place. Language in Society 2(1): 45–80.

19.

Lakoff

(1975) Language and Woman’s Place. New York: Harper & Row.

20.

Lakoff

(2004) Language and Woman’s Place: Text and Commentaries ( Bucholtz

, ed.). Oxford: Oxford University Press.

21.

Macalister

(2011) Flower-girl and bugler-boy no more: Changing gender representation in writing for children. Corpora 6: 25–44.

22.

McEntee-Atalianis

(2013) Stance and metaphor: Mapping changing representations of (organizational) identity. Discourse & Communication 7(3): 319–340.

23.

Motschenbacher

(2013) Gentlemen before ladies? A corpus-based study of conjunct order in personal binomials. Journal of English Linguistics 41(3): 212–242.

24.

Partington

Duguid

Taylor

(2013) Patterns and Meanings in Discourse: Theory and Practice in Corpus-Assisted Discourse Studies. Amsterdam: John Benjamins.

25.

Pearce

(2008) Investigating the collocational behaviour of man and woman in the BNC using Sketch Engine1. Corpora 3: 1–29.

26.

Scott

(2017) WordSmith Tools version 7. Stroud: Lexical Analysis Software.

27.

Tannen

(1994) Gender and Discourse. New York, NY: Oxford University Press.

28.

Taylor

(2013) Searching for similarity using corpus-assisted discourse studies. Corpora 8: 81–113.

29.

United Nations General Assembly (2015) Transforming our World: The 2030 Agenda for Sustainable Development. Available at: http://www.un.org/ga/search/view_doc.asp?symbol=A/RES/70/1&Lang=E (accessed 27 May 2020).

30.

UN Women (2016) Status of Women in the United Nations System. Available at: https://www.unwomen.org/en/digital-library/publications/2016/12/status-of-women-in-the-united-nations-system (accessed 27 May 2020).

31.

Weatherall

(2002) Gender, Language and Discourse. London: Routledge.

32.

Wodak

(2003) Multiple identities: The role of female parliamentarians in the EU Parliament. In: Holmes

Meyerhoff

(eds.) The Handbook of Language and Gender. London: Blackwell, pp.671–699.