5. Fact-Checking Language

This chapter is about how to fact-check words. As a fact checker, you will not always be involved in conversations about the language used in a story, nor should you copy-edit or proofread the story you're working on. Your role is to speak up only in cases where word choice or punctuation concerns factual correctness. In brief, you are responsible for checking any diction, spelling, or grammar that changes the meanings of statements—and therefore the facts about the world that they convey.

Language use and accuracy are closely related. Accurately portraying an event or someone's experience is as much about the facts described as it is about the words used to describe them; linguistic precision is an important tool for conveying someone's personal identity and lived experience. Just as importantly, some words and phrases can reinforce inaccurate stereotypes and harmful narratives about people or signal a political stance. As journalists, our responsibility is to ensure that the language we use in fact-checked journalism—the record we create—correctly represents the facts, including inclusive linguistic norms.

We begin this chapter by discussing the importance of accurate language in fact-checked journalism and distinguishing between the roles of fact checker and copy editor. Then, in Five Important Types of Language, we list a number of situations in which language is a fact checker's responsibility, and afterwards, we explain how fact checkers can approach such situations rigorously and responsibly by working with appropriate references and interviewing people about language.

The Importance of Accurate Language

Language is a fundamental tool in journalism; it is the primary medium in which facts are communicated. This means you should be conscious of language in your evaluations of and communications with sources as well as in the text of the story you're fact-checking.

As mentioned in Chapter Four, assessing the language used in a printed source can help you determine its authority with respect to certain facts. And, as we will discuss in Chapter Seven, the language you use when communicating with sources can affect their comfort level and whether they feel respected by the fact-checking process.

Fact checkers need to know about appropriate language use in different circumstances. But they also sometimes need to apply their verification methodology to language, especially for questions of accuracy regarding punctuation, spelling, style, and word choice. In the following sections, we will discuss all of these, but we will pay particular attention to word choice. In many communities, language is a way of establishing trust and showing respect. Using the wrong words can be not only inaccurate but also harmful and alienating. Using person-first language to describe someone living with a disability, for example, is both accurate and respectful—unless the person you're describing prefers different language, which you usually won't know until you ask them.

It is the fact checker's responsibility to be aware of the factual basis for each word included in a story, which means they need to understand what each word means. Precision is key: you should know what grounds the truth of a statement and which words in it can and can't be replaced. This is not something you can always determine by opening up a dictionary. It is sometimes also an issue of connotation and context, since certain words have shifted meaning over time or become the subjects of political debate. Other times, it is a question of authority: dictionaries don't always include the words, spellings, and definitions that are standard in different communities. Still other times, it is an issue of linguistic self-determination: in some contexts (though not all), an individual, community, or social group gets to decide for themselves what language is used to describe them.

The relationship between language and identity is especially important: gender, sexuality, race, and other aspects of personal identity are all mediated and communicated in part through language. According to Craig Silverman, a journalist who wrote a book on corrections, misidentification is one of the most common media errors. So fact checkers need to pay particular attention to the language used by reporters to describe people.

Finally, fact checkers should understand that language is often used as a weapon: people argue about words and ascribe moral values to them, and they use language to include or exclude others. It's sometimes impossible for a journalist to stay "neutral" or "objective" about such issues: we can't avoid using language. Your objective is not to "rise above" these linguistic issues but rather to understand them well enough to navigate them. We break this skill down into three parts: identifying your domain, choosing your references, and asking questions.

Fact checkers are not expected to right the world's linguistic wrongs (nor are copy editors, though they may have a little more responsibility here). But we are responsible for identifying the situations in which factual issues of language may come up and making sure that the appropriate conversations have taken place before certain language is used in a story.

Usually, our job as fact checkers is to confirm the accuracy of a description of a source—if the writer has put down that they have brown hair and are tall, are both of those things true?—but sometimes the way someone is described by a writer isn't something we should check as neutrally as any other detail. As an example, if you are checking a story that describes someone as "fat," that's a moment when it might be worthwhile to pause. "Being fat" is something that can carry stigma in our society, and many people wouldn't want to be publicly described this way.

On the other hand, many people identify as fat and work to reduce both the stigma of that description and the barriers, such as medical discrimination, that fat people face. The word fat itself isn't inherently unkind, and many people would embrace that label. But you have to exercise judgment when you're checking. There are many harmful and hurtful stereotypes about fat people—Is the writer using the physical descriptor as a shorthand for something about that person's character? Is the person's appearance relevant to the story? It may be that the description is welcomed and appropriate, in context, but the point is that the words writers may use to describe someone aren't value-neutral, even if they seem superficially easy to check.

—Erin Sylvester, Managing Editor, The Walrus

It's a well-established problem in the terrorism literature that there's no single, agreed-upon definition of the word terrorism...I'm working on a newsroom study that involves talking to editors and journalists at legacy publications across the country. And one thing I'm asking them about is whether their newsrooms have any guidelines regarding when to use the term terrorism and when not to. So far, no one has answered yes to that question. Most newsrooms don't seem to have rules or guidelines in place—it's at each individual's discretion. And that can be a huge problem...

I've found in my research that the term is applied disproportionately to Muslim subjects, because the use of the word terrorism in American news media is informed by decades of immersive, insidious Islamophobic coverage of Muslims. And I don't necessarily think this is a problem that can be traced back to just individual bigotry or bias, although bigotry and bias are real problems. This is a systemic problem informed by big-picture institutional biases.

—Ruth DeFoster, Assistant Professor, Hubbard School of Journalism and Mass Communication, University of Minnesota

Working with the Copy Editor

Most publications work with a copy editor, though of course this is not always the case. (As mentioned at the end of Chapter Three, some publications also hire one person to be both fact checker and copy editor, but this working arrangement is far from ideal.)

As part of a standard editorial team, the copy editor is the person responsible for ensuring consistency within a story and across different stories in the same publication. They do so by (maybe) building and (definitely) maintaining and updating a "house style" and making sure that editorial copy reflects it.

The division of responsibilities between fact checker and copy editor essentially boils down to the division between matters decided by house style and matters decided by "the facts." A copy editor should verify spelling, word choice, punctuation, and other style-related issues that change nothing about what the statement in question conveys. They are concerned, for example, with whether an anglophone publication uses British or American spellings, whether they hyphenate fact checker or fact-checking, and whether they use quotation marks or italics when referring explicitly to a word. But, in many cases, copy editors can't make informed stylistic or grammatical decisions about a story without some collaboration with the fact checker.

Take the sentence: For Jule, a queer graduate student of colour in their mid-twenties, attending the student protest was "both inspirational and overwhelming": the situation on campus was clearly about to change for the better, they thought, but it wouldn't be an easy or efficient transition.

The following questions need to be answered:

1. Do the statement's punctuation and spelling match house style?

2. Are the facts the statement means to convey accurate?

3. Do the words in the statement mean what the reporter or interviewed source (Jule) thinks they mean?

4. Is the language appropriate in the context of the story?

5. Do the words associated with a source represent that source accurately?

The copy editor is solely responsible for answering question 1, and the fact checker is solely responsible for answering question 2. But questions 3 to 5 require collaboration between the copy editor and the fact checker, who will need to determine, for example, the spelling of Jule's name, whether Jule identifies as a queer person of colour (and does not prefer another descriptor), whether Jule would describe the student protest as "inspirational and overwhelming," what those words mean to Jule, and whether Jule thought it was "clear" that the situation on campus would soon improve. The copy editor decides whether to spell it color or colour, but the fact checker decides whether person of colour is an accurate descriptor in the first place.

According to the standard editorial process (see Chapter Two), a copy editor works on a story only after fact-checking is complete. The copy editor therefore presumes that the fact checker has already done their job and that, if anything factual comes up during copy editing, they'll be able to get the answers they need from the fact checker.

In some cases, grammar will affect facts directly, and the copy editor will ask the fact checker about the placement of punctuation or phrases. For example, the statement Viviane's sister, Eloïse, studies psychology is factually different from Viviane's sister Eloïse studies psychology: the commas imply that Eloïse is Viviane's only sister.

The statement Fresh from the oven, Ally offered Viviane another baguette is factually incorrect due to its misplaced modifier (unless Ally just came out of the oven).

And the statement "I don't like the texture of soft cheese," Ally says is factually different from the statement "I don't like the texture of soft cheese," Ally said. The tense difference in the verb to say is key: the first version implies that Ally still believes these words—meaning that she likely spoke them during an interview with the reporter—whereas the second implies that she once spoke those words, for example in a press conference, but may not stand by them.

These distinctions reflect grammatical nuances that are often overlooked by reporters (and everyone else). It's the copy editor's job to correct errors and identify the right phrasing, but they won't be able to do so in these cases without consulting with the fact checker about the facts. The same is true for the spellings of proper names, locations, acronyms, and other terms related to the external world (for example, is the bakery where Ally and Viviane bought baguettes called The Bakers' Corner or The Baker's Corner?). Fact checkers should anticipate questions regarding spelling, punctuation, and style and be ready to answer them; you might even leave notes in your checked copy so that they can be seen by the handling editor during fact-check input and relayed to the copy editor afterward.

If ever you're unsure whether a linguistic issue is within your purview, it's a good idea to consult with the copy editor. They can guide you, along with the head of research when appropriate, toward asking the right questions about language during your fact-check.

If, over the course of your fact-checking, you determine that some word choices are crucial for accuracy—for example, if the story's central character identifies as a survivor of sexual assault but not as a victim—make sure the entire editorial team is aware of this by noting it explicitly in your checked copy, regardless of whether the current language requires a correction. That way, your colleagues will understand the word choice's importance and know not to modify it during the editorial production process. If the wording goes against house style, the copy editor has final say on whether it can stay in the story, but that decision will likely be the product of conversation among members of the entire team (including you), taking into account the facts you've confirmed.

How to Fact-Check Language

Here is a list of the language-related facts you will generally be expected to correct and be ready to discuss with the copy editor:

• Punctuation: anything that affects the facts, including the placement of commas and where quotation marks should begin and end to demarcate a source's quote.

• Abbreviations, initialisms, and acronyms: for example, whether NASA is spelled correctly and whether the name is an initialism (spelled out in common parlance) or an acronym (spoken as a word), which affects whether it needs a definite article (NASA, but the RCMP).

• Proper names: including the names of cities, neighbourhoods, and companies.

• Pronouns, honorifics, and titles: for example, which pronouns someone uses or, if they are called "Dr.," whether this is a fact about their medical training or their academic credentials.

• Word placement: anything that affects the facts, for example the difference between She held the alleged murder weapon and She allegedly held the murder weapon.

• Modifier placement: anything that affects the facts, for example, the misplaced modifier in Fresh from the oven, Ally offered Viviane another baguette.

• Ambiguity and vagueness: for example, if a scientific study on fertility was conducted on a sample of "young adults," what ages count as young?

• Technical language: including the proper use of terminology from different research domains and industries, for example the measurement units used by physicists.

• Statistical language: for example, whether the decrease in someone's annual carbon consumption was 5 percent or 5 percentage points (see the guidelines on data and statistics in Chapter Three.

• Foreign words and translations: including whether words have been spelled or translated correctly, especially if this work was done by the reporter. (This may require collaborating with someone who is familiar with the language in question.)

• Direct textual quotations: whether the reporter has cited a book, article, or other written work accurately (including the original spelling and punctuation) within a quotation.

• Context-dependent meanings: any terms whose meanings may shift depending on the context in which they're presented, for example the word soda in different regions of North America.

• Cultural terms: including words that may not appear in a dictionary but are part of a cultural vernacular, as well as alternative spellings for words that are in the dictionary.

• Polarizing language: terms with moral or political connotations, such as known to police or global warming.

• Language describing personal identity and experiences: whether the terms and phrases would sound accurate to the person whose experiences and identity are being described.

• Language describing social identity and community experiences: whether the terms and phrases would sound accurate to the community whose experiences and identities are being described.

Everything in this list is important, but where things get most complicated for fact checkers is questions of word choice, especially related to the last five items: context-dependent meanings, cultural terms, polarizing language, and personal and social identity.

In most domains listed above, you won't have to call anyone to fact-check the language involved; referring to gathered sources should be enough. (You do not need to speak with the authors of a book to check that quotes from their book are accurate, for example, as long as you have access to a copy.) However, in the last five cases, you likely will need to speak with people to confirm the accuracy of the reporter's language. In the rest of this chapter, we will focus largely on this responsibility.

One particularly useful tool in this context is the Internal/External Fact Distinction, which allows us to separate statements describing someone's identity and experience into their internal and external elements. Fact-checking language is often an internal question, but not always—it depends on what the language is meant to represent and which person or community we take it to be internal to.

Five Important Types of Language

In general, five types of word choice cause the most difficulty for fact checkers: context-dependent meanings, cultural terms, polarizing language, personal identity and experiences, and social identity and experiences. Of course, these categorizations are rough, and in practice, they often overlap. Cultural terms, for example, may have context-dependent meanings, and someone's identity may not be clearly separable into its personal and social elements, particularly if it is articulated in cultural terms. But this categorization can, at the very least, help you get started in your fact-check by identifying the kinds of sources you should be looking for to check the different types of language facts. Once you understand the different strategies required by different categories, you'll find it more helpful to consider their overlaps and plan accordingly.

Below, we outline the types and strategies in detail, and in the following sections, we explain how to fact-check with the appropriate sources, including dictionaries, community style guides, and people.

• Context-Dependent Meanings

  The meaning of some words depends on the context in which the word is used; these words have context-dependent meanings. The fact checker is responsible for determining the context in which a word is being used in a story, and so they are essential to determining its correct meaning.

  There are, for example, regional differences in the meanings of English words; if someone is described as drinking "soda," you will need to account for the regional context in which the description was used and, depending on the answer, check whether they were drinking carbonated water or a sugary drink. If someone was said to have drunk five "pints" of beer at the bar, then you'll need to check whether those were British or American pints (20 or 16 ounces), if that detail is relevant to the story.

  Technical terms are another typical example of context-dependent meanings. To give a simple example, the meaning of winter will depend on whether the statement is being made in the context of the Northern or Southern Hemisphere. And the dates winter refers to will depend on whether the reporter or source is using the astronomical or meteorological definitions of the seasons.

• Cultural Terms

  Words travel often between languages and cultures—they are appropriated, eventually taken out of context, and given new meaning, all as part of language's natural evolution. Still, there is the ethical question of when journalists should avoid using certain words that have been appropriated, such as powwow. This is an important topic, but it belongs largely in the copy editor's domain. (Nothing prevents you from bringing up offensive language when you see it in a story—it's never a bad idea—but your official purview is the facts.)

  Fact-checking and cultural language usually intersect when the reporter attempts to use a cultural term (or a word in a foreign language) as it is used by the community in question. For example, in a story about a city's local Polish restaurants, you will have to verify that the words grochówka and kielbasa are spelled correctly and used accurately, even if only the latter is currently included in your publication's house dictionary. And, in a story about Canadian slang, you'll have to confirm the same about darts (cigarettes) and dep (corner store, from the French dépanneur), even though neither word is included in the dictionary.

  The names of places and languages are also sometimes part of this category—they can vary based on the person, community, or culture. For example, there are different accepted variant spellings of the language name Ojibwe (Ojibwa, Ojibway, etc.), and the best one to use in a story may depend on the source, person, or community being referenced. (We will discuss this further in Working with Language and Style References.)

• Polarizing Language

  Certain words become "loaded": the ways they have been used give them additional connotations beyond what is included in their formal definitions. This is what we call polarizing language. One example of this is the term known to police. Technically, the term means that the police have a record of someone—because, for example, that person called the police for help, they were the subject of a complaint, or they were arrested. The language itself should be morally neutral: many people are "known" to police for all sorts of reasons. But the term carries a negative connotation because it is often used to describe repeat offenders or illegal activity. So, while the explicit fact conveyed by the use of such a term in a story may be accurate, the implicit fact may not be. You need to be aware of these connotations so that you can check them—and correct them, when necessary.

  Some polarizing language is the subject of intense debate, and it can be difficult to determine which words to use—though one choice may be more accurate than another. Terms for gender and sexual identity are often politicized and require careful fact-checking. A reporter using the term opposite gender, for example, implies that they do not understand that there are more than two genders, and the perpetuation of this misunderstanding in the media is a serious issue.

  It is sometimes impossible to stay politically "neutral" with respect to language choice. For example, your publication will have to decide what words to use to describe the global climate crisis. Scientists and prominent newsrooms have advised calling it a climate emergency; whether or not you choose to adopt that language can be seen as taking a particular stance regarding the facts of climate change. In these cases, the best you can do is discuss the issue with the editorial team (and any relevant sources in the story), take a reasoned and fact-based stance on the matter, and be consistent.

• Language describing personal identity and experiences

  Language plays an important role in any identity statement. If someone's identity is described by the reporter, there are multiple facts you may have to confirm: whether they do identify that way, whether the social group to which they claim to belong would agree with the language being used to describe that identity, and whether the external facts related to that identity statement are true. That first element—whether someone identifies a certain way—is an internal linguistic fact: it can be confirmed or refuted only by the person in question. This personal identity language concerns only the person it describes, so it is accurate to the extent that they feel it is accurate. The second element is social identity language, which we will address in the next bullet point, and the third is concerned with external facts about identity, which we will address in Chapter Six.

  For some statements of personal identity and experience, there are no external facts to check: the fact that the person considers the language correct means that it is correct. For example, there is no "fact of the matter" about someone's pronouns beyond the fact of which pronouns they use and identify with—their language cannot be "proven wrong" by the external world. Another example is descriptions of how someone is feeling: the only source that can confirm such a fact is the person who felt the feelings, and they get to decide which language sounds right to them. This means that, even if you have a reliable secondary source describing someone's identity or feelings (such as a fact-checked magazine profile of them), it's always better to speak with the person directly during fact-checking to confirm those statements.

  But there are cases where the multiple layers of identity language can come apart. For example, although there may be an external fact to check about someone's relationship to alcohol, it is an internal fact which words they use to describe that relationship. Do they identify as someone with an addiction, someone with a substance use disorder, someone recovering from alcohol addiction, or otherwise? That's a question of personal identity language, which only they can answer. Using other sources, you can also check the related external fact of whether they are currently attending regular AA meetings.

  The same idea applies to whether someone identifies as a victim or survivor of sexual abuse: their choice of words is an internal fact, but the experience described by those words is at least partly an external fact.

  Sometimes, the correctness of personal identity language will depend on whether it is being used in a source's quote or in the reporter's main text. Within a quote, you should always use the language the source prefers to use to describe themselves; outside of a quote (in the reporter's voice), you may choose to use other language in certain circumstances. As we mention below, if the term someone identifies with goes against their community's established norms, you may have to make that clear in the story.

• Language describing social identity and community experiences

  Just like individuals, communities and social groups often use language as a tool for self-determination. Marginalized communities are typically very aware of the language that is used to describe and define their identities and experiences, and they often have established opinions about the appropriateness of certain terms. The resulting norms typically change over time, and it is important to be aware of current language preferences while working on a story. Since the fact-checker will be the one speaking with the sources in a story, they will be responsible for discussing language preferences with sources, including any divergence from community norms. This is the purview of social identity language.

  Just as personal identity language reflects how a person choose to describe themselves, social identity language reflects how a social group or community chooses to describe itself.

  Style guides such as the National Association of Black Journalists Style Guide, for example, can teach mainstream media about the inaccuracy of a previous word use and provide current definitions of certain culturally relevant words. One person may identify as African American, and another may identify as Black. Both of these are acceptable according to the guide—whereas Afro-American is not.

  Not all communities will have set preferences for how their members should be described, so you should be aware of the context in which different terms are used. There may even be serious disagreement within communities about language use—for example, the descriptor autistic person versus person with autism. As mentioned in the previous section, you should always ask for interviewed sources' personal preferences about which language is used to describe them.

  Note that personal and social identity language do not always match. Someone who willingly works in the sex trade may identify as a prostitute, for example, though the generally accepted community term is sex worker. The general guidance is that the language describing an individual in a story should reflect the language they use to describe themselves; that means, in this case, you could include prostitute as part of the source's quote, and you should always check with sources about their own feelings regarding their community's preferred language during the fact-checking call. But, when the story is told in the reporter's voice, the language should generally follow community norms; in this case, the reporter should use the term sex worker. (If you are fact-checking a memoir for which the author has decided to use language that goes against community norms, you can suggest including a note explaining the decision within the text of the story itself.)

  Not all Black people are African Americans (if they were born outside of the United States). Let a subject's preference determine which term to use. In a story in which race is relevant and there is no stated preference for an individual or individuals, use [Black] because it is an accurate description of race. Be as specific as possible in honoring preferences, as in Haitian American, Jamaican American or (for a non-U.S. citizen living in the United States) Jamaican living in America. Do not use race in a police description unless the report is highly detailed and gives more than just the [person's] skin color.

  —National Association of Black Journalists Style Guide

It's easy to see how often these categories will overlap. Reclaimed words, for example, can be used appropriately only when spoken by members of certain communities in certain contexts. Queer is a reclaimed term that is part of social identity language and that has been given context-dependent meaning. Some members of the queer community use it often, while others still consider it to be offensive, so the term's context is important to check.

Social identity language is often politically salient. The Trans Journalists Association's Style Guide, for example, explains the difference between the terms genderfluid and gender nonconforming: the latter describes someone living outside typical gendered expectations while the former describes someone whose gender is not fixed. These linguistic details make a difference to the facts conveyed in the story—and they concern a number of the categories listed here (personal identity language, social identity language, cultural terms, and polarizing language). A fact checker may need to confirm whether someone identifies as gender nonconforming, and that requires first understanding what the term usually means according gender nonconforming people, then checking the fact with them using that definition.

Language and identity are messy and full of grey areas. Our understanding of personal and social identity and the language we use to describe them is constantly evolving. As a result, no set of guidelines will ever account for every situation you may encounter during fact-checking. But understanding how to navigate the kinds of sources you may use to check different linguistic facts is a good start—and, in the next section, we will help you do just that.

Working with Language and Style References

Dictionaries are, of course, a journalist's reference of choice for all matters concerning words. The publication for whom you are fact-checking likely has a house style guide as well as a default dictionary—meaning that the copy editor has already decided which authorities to defer to for questions of spelling. These are the references you should turn to if you don't know what a word means when fact-checking a story.

Dictionaries are a helpful first step in fact-checking research: they allow you to inform yourself about the appropriateness of certain terms before you conduct further research. They can also teach you how to pronounce or paraphrase words ahead of a fact-checking call, and they can alert you to when factual clarification may be needed. But that doesn't mean the information in a dictionary is always authoritative: the dictionary is not the final word on words. In fact, there are times when it's best to set aside your dictionary, especially in the scenarios listed in the previous section.

According to Kory Stamper, a freelance lexicographer, a dictionary is a "snapshot of a part of a language that was taken at one time by a particular group of people at a particular place." In America, Canada, and the United Kingdom, that snapshot tends to be from the perspective of "a group of highly educated white people." A dictionary cannot account for all of the variations and evolutions of language across different regions and times. (This is especially true for print dictionaries, which are updated less frequently than online ones.)

Robert Jago, a journalist from the Kwantlen First Nation, gives an example of the limitations of dictionaries for fact-checking cultural terms. Imagine that a reporter from the Kwantlen First Nation writes a story that includes the words mitzvah (in the sense of a Jewish commandment) and blanketing (in the sense of an Indigenous blanketing ceremony). The copy editor would check that these words are spelled correctly and decide whether to italicize them or define them in the text. But, since these are cultural words, it would also be the fact checker's job to confirm that they accurately describe the events the reporter uses them to describe. And the fact checker's first stop would be the dictionary.

However, if you refer to the Merriam-Webster online dictionary, for example, you will find the intended use of mitzvah but not that of blanketing. If you were assigned that story to fact-check, you would be able to verify that the word mitzvah exists and is being used correctly. But how could you check that the term blanketing exists and is being used correctly? To start, by broadening the sources you use and turning to references from within the culture that uses the word. (We would recommend looking for an Indigenous source to confirm the word's spelling and use, and perhaps speaking with someone from a First Nation in which blanketing is practised—in this case, another member of the Kwantlen First Nation, perhaps someone working in their cultural affairs office or band museum, if those exist.)

Anglophone dictionaries are particularly deficient when it comes to terms related to Indigenous cultures. Knowing whether to use the word reserve to describe a First Nations community is something you won't be able to determine by looking at the Oxford English Dictionary, for example, nor will you find the current preferred spelling for the names of different nations and tribes in the Canadian Oxford Dictionary (which, as of this writing, hasn't released a new version since 2004 and still capitalizes the word Internet).

Dictionaries also have other deficiencies. For one, they likely aren't current enough to keep up with social and community-based linguistic norms, which are related to personal identity language and social identity language. And they likely aren't up to date on the significance of polarizing language. Some technical terms can be found only in specialized books and dictionaries, not all of which may be accessible to you. So, while dictionaries can help with standard, uncontroversial terms, they likely won't be helpful to your fact-checking precisely when you need them most.

Dictionaries are really good at denotation of meaning. Like what's the definition of homosexual, that's really easy. But they've not been really good at registering connotation of meaning, which is to say, the noun homosexual is outdated and it has medical stigma attached to it. So do not use it to describe people unless they self-identify as that.

—Kory Stamper, freelance lexicographer (Merriam-Webster, Cambridge Dictionaries)

As mentioned in the previous section, community style guides are helpful for filling in the gaps left by dictionaries. Some style guides are broad in scope, such as the Conscious Style Guide, while others are more focused, such as femifesto's Use the Right Words for reporting on sexual violence and Gregory Younging's Elements of Indigenous Style: A Guide for Writing By and About Indigenous Peoples. These guides have been created for the purpose outlined above: to guide editorial teams when dictionaries don't give the full contemporary picture. We recommend deferring to these style guides over dictionaries most of the time and, when appropriate, even discussing their use with the sources quoted in a story, since they may have preferences or suggestions regarding style guides. But this is not entirely in your control—the copy editor has considerable input on all issues of style.

The whole idea of dictionaries being this huge authority in language and knowledge, it's a marketing tactic that dictionaries started back in the 1860s to sell dictionaries. And it's not true. There are a lot of dictionaries with a lot of stuff in them that's really objectionable and really wrong. And that's because they are a product of their time and their culture, and their culture is a very specific culture.

—Kory Stamper, freelance lexicographer (Merriam-Webster, Cambridge Dictionaries)

It's worth noting that some political conflicts about words include the use of pseudo style guides released by organizations that aim to block certain progressive or inclusive language. Some trans-exclusionary "feminist" organizations, for example, which we will not link to here, have released style guides advocating for the use of words that exclude and denigrate trans people. (Helpful tips for identifying trans-exclusionary rhetoric can be found here.) Before using a community style guide, you should always make sure to look at who made it and what their motivations are.

Finally, on some occasions, you may not have a style guide to fall back on. We often encounter this when reporting on minoritized communities, where language revitalization and/or linguistic self-determination may still be ongoing or in early stages, and regional variation may not have been standardized. For example, there is no universal agreement about the spelling or use of certain names and terms in many Indigenous cultures (largely due to the attempted erasure of those cultures by colonial forces over the past several centuries). Some Indigenous communities that are focused on linguistic revitalization have created their own dictionaries, and different dictionaries often include different spellings for the same term.

In these cases, people are your best references for fact-checking words. If the reporter is quoting someone who is using a word that isn't in any dictionary or style guide, you can call them to ask how they spell it and make sure that spelling is reflected in the story. If their spelling conflicts with the spelling offered by a community dictionary or by another source who is quoted in the same story, you can ask them whether they have a preference. For the most part, it's preferable to include two spellings in a story (and explain why that is the case) than to include one that does not accurately reflect the sources and the state of the language at the time.

If you've got three different spellings of an Indigenous name or even a personal name, then I think it's on the publication—particularly if you are a majority-language publication representing a minoritized group or language—to help carry forward that knowledge in a way that is accessible to majority audiences but also honours minoritized people in the choices that they've made.

—Kory Stamper, freelance lexicographer (Merriam-Webster, Cambridge Dictionaries)

We chose not to standardize the spelling of Indigenous words or names in this issue. We want to preserve regional dialects, and we understand that many Indigenous words are artificially standardized by colonial researchers. For example, "Anishnaabe," "Anishinabe," "Nishnawbe," and "Neshnabé" are all terms that refer to the same group of people, but have different spellings in different dialects. Writers were free to use whatever spelling they thought was most accurate, or whatever spelling they arrived at in consultation with members of the community.

—"'Land Back' is more than the sum of its parts: Letter from the Land Back editorial collective," The Land Back issue, Briarpatch (2020)

Instructions for Identifying the Appropriate Language Reference

• Is the term included in the dictionary, and does the dictionary assign it the spelling and meaning one would reasonably expect? If so, you may use the dictionary.

• If not, does the word relate to a culture, community, or practice? If so, is it an instance of:

  • Context-dependent meaning or technical terms? Then it should be checked with context-appropriate technical references or subject-matter experts.

  • Cultural terms? Then it should be checked with community style guides or representatives.

  • Social identity language? Then it should be checked with community style guides or representatives.

• Regardless of your answers above, does the word also describe an individual? If not, your work is likely done. If it does, speak with that person to confirm the internal linguistic fact.

• If there is conflict among these priorities, think critically about the placement of quotation marks, which source should be considered most authoritative for the fact at hand, and what proper acknowledgement of that conflict in the piece would look like.

Interviewing Sources about Language

The language fact checkers use matters, especially when speaking with the interviewed sources in a story. We will not go through all the details of a fact-checking interview here—for general instructions, see Chapter Three—but a few points are important to keep in mind.

First, you will need to conduct some background research (using dictionaries and style guides, for the most part) to ensure that the words you use in your fact-checking questions allow you to actually check what you think you're checking. Make sure that you know which words need to be fact-checked verbatim with a source (such as identity and other internal linguistic facts) and which can be paraphrased because they do not hinge on word choice. Do not hesitate to ask follow-up questions to ensure that the source meant what the reporter thought they meant by using a certain term or phrase. (To that end, make sure that you know what the reporter themselves meant to say. If you aren't sure, reach out to the reporter and ask clarifying questions before reaching out to the source.)

Be transparent with sources about your work and why you are asking certain questions about language. When looking into cultural spelling differences, for example, explain why you are asking to confirm the spelling (including the fact that you couldn't find an official spelling in a dictionary or style guide during your initial research) and explain the significance of setting a particular spelling or definition down in a fact-checked journalistic record.

Finally, be conscious of the way that you phrase questions about language and identity. Kate Sloan, a freelance journalist who focuses on sex and relationships, recommends wording fact-checking questions about language in a way that makes the description component explicit.

Language in the queer community is ever evolving, and it's a huge way people communicate competency. If a journalist is using outdated language, that can put a wall up for the source; as a journalist, it's very important to stay up to date with that language.

Especially in the realm of gender, but also in the realm of sexuality, there has been a push to stop using the phrase "identifies as" when describing someone's personal labels because it can imply that the speaker thinks the subject's identity is somehow false or otherwise suspect. Being queer, trans or non-binary is not just about labelling oneself these things; those labels are reflective of who/what someone is, and these identities are often considered to be inborn and fixed; to say that someone "identifies as" something that they are is in some sense diminishing that. So, for instance, rather than asking someone, "Do you identify as non-binary?" I might instead ask any of the following: "Are you non-binary?" "Is 'non-binary' a word you would use to describe yourself?" "What words or labels do you use to describe your gender?"

—Kate Sloan, freelance journalist

Sources have little to no motivation to be dishonest about who they are, what their ethnic or racial backgrounds are, or what their gender identity and sexual orientation is. It is appropriate for us as editors to ask our sources to confirm how they identified themselves to ensure that we are being accurate in our reported pieces.

—Lara Witt, Editorial Director, Prism

When it comes to social identity, a similar approach can be taken, though fact checkers should be conscious of their limitations when it comes to arbitrating the external facts of social identity (see Chapter Eight). Do not be overly skeptical unless red flags come up—the point of fact-checking identity language with an interviewed source is simply to check that the language being used is accurate, not necessarily that the identity description itself is correct in a broader sense.

Sloan's recommended framing is similarly applicable here: instead of asking "Do you identify as Latina?" you might ask, "Is Latina a word you use to describe yourself?" Jackie Wong, a senior editor at The Tyee, also suggests asking, "Do you describe yourself as belonging to this community?" (We will discuss how to actually fact-check external facts about social identity in next chapter.)

When a fact checker is checking a point about someone's social identity, I think it's helpful to provide as much choice as possible. Some folks are passionate about sharing the intersections of their identity, which they feel are key to who they are, and others are not; in my opinion, there should be room for both.

—Jackie Wong, Senior Editor, The Tyee