Tell the story: How to write for American Ethnologist

Open access article


Read Article


To write a successful article for a major anthropology journal, authors can employ a number of strategies. The first and most essential is to familiarize themselves with the journal's mission. For AE, authors must ground their arguments in current debates in the discipline, make a clear contribution to anthropological theory, support this contribution with ethnography, and demonstrate its relevance to contemporary sociocultural problems. Moreover, AE authors must cite other authors appropriately, cohesively structure the article, and avoid various problems that afflict much of academic writing—such as evasive metadiscourse, inflated diction, serial citation, and unnecessary literature reviews. To maximize the article's online discoverability, authors should judiciously compose their titles and abstracts and carefully choose their accompanying keywords.

Since 2015 we have been working as, respectively, editor‐in‐chief and senior copyeditor of American Ethnologist. During this time the first author has read, evaluated, and commented on hundreds of manuscripts and read, interpreted, and distilled thousands of reviewers’ reports. The second author has worked with dozens of authors to copyedit and polish their accepted manuscripts. It now strikes us as a propitious moment to draw on this experience and offer some advice to prospective authors, especially those in the early stages of their scholarly careers. What makes a successful submission to AE and, more broadly, to a general‐interest anthropology journal?

This question is seldom addressed in our discipline, even though it is shrouded in mystery in the minds of many scholars. This may explain the exceptional popularity of a two‐hour workshop on journal publishing that Niko Besnier, since becoming editor of AE, has given at several anthropology departments and conferences around the world (in four languages so far). Much of the material we present here is based on this workshop. By making this material widely available, we hope that prospective authors will improve the quality of their submissions to AE and comparable journals, making their work more likely to be accepted for publication.

We are not the first to write about submitting to anthropology journals. For example, former American Anthropologist editor Tom Boellstorff (2008, 2010, 2011; Vora and Boellstorff 2012), as well as others (e.g., Holbraad, n.d.; Unger 2017), offers sound advice on the topic, and we encourage readers to consult these useful sources. [1] Among other things, they advise authors to ensure that their data support their theoretical claims, to avoid sweeping generalizations, to be attentive to presentation (avoiding typos, eliminating tracked changes, etc.), and to be polite when corresponding with editors, copyeditors, and editorial assistants. We agree wholeheartedly with all these suggestions (particularly the last one).

Here, however, we aim to do something different: to offer advice on what makes a successful manuscript, from broad questions of how to connect theory with ethnographic data to more specific problems of composition. Along the way we will also point out issues that usually signal that the manuscript needs more work. We do not aim to provide a universal set of guidelines on how to structure arguments and construct an effective text—an impossible task, since there are so many ways of doing so. Nor do we present an exhaustive treatise on every aspect of writing good scholarly prose. Rather, the following discussion focuses on key problems in the manuscripts we review, problems that appear so often that we feel compelled to address them.

American Ethnologist’s mission

Before submitting an article, an author must ensure that it is appropriate for the journal in question. For early‐career scholars in particular, choosing which journal to submit to can be a formidable task for which few guidelines are available. Determining the submission's “fit” is an important task, although authors must bear in mind that fit is not the only factor in editorial decisions. One way to determine whether a journal is a good choice is to carefully read recent issues; a surprising number of authors seem to skip this basic first step. The author should focus on the kinds of themes the journal emphasizes, the balance between abstractions and empirical materials, and the journal's theoretical ambitions. The author should also pay attention to the textual organization of articles, such as section structures (e.g., are they numbered?), endnotes (how many in a typical article?), and style (how formal?), since the more closely a submission resembles published articles, the more likely the editor and reviewers are to see it as viable.

The other essential step is to carefully read the journal's website (many potential authors also seem to skip this step, even though at submission to AE they are asked to confirm that they have “followed the journal's instructions on the form and style of the title, abstract, keywords, main text, notes, and references”). One central item on the website is the journal's mission statement. “American Ethnologist,” it declares, “is a quarterly journal concerned with social and cultural anthropology in the broadest sense of the term.” It continues, “The journal's articles combine ethnographic specificity with original theoretical thinking, conveying the relevance of the ethnographic imagination to the contemporary world.” The key terms in these statements are “theory,” “ethnography,” and “relevance to the contemporary,” which capture the three main aspects of the journal's mission.

Three keywords

Not all anthropology journals have the same aims and standards. Some specialize in a specific area of knowledge or a particular region of the world. But even journals that address generalist readers differ from one another. While most editors are prepared to consider manuscripts that fall outside the journal's traditional scope, they also have a responsibility to maintain a certain orientation that their predecessors have established and that has contributed over the years to the journal's “branding.”


Theory can mean different things in different contexts, and anthropologists often disagree about what is theoretically important. What is considered theoretically significant varies, for example, according to different national traditions of anthropological inquiry. Theory in our discipline also changes over time, not only as particular theoretical approaches wax and wane but also as scholars redefine what needs to be explained or abstracted in human life, and how. It is often said that the days are long gone when anthropology was dominated by great theoretical debates (e.g., structuralism vs. Marxism vs. symbolic anthropology), leaving a field in which theory is eclectically cobbled together, often with ideas emanating from outside the discipline. Furthermore, some theoretical approaches have longer shelf lives than others: while some are little more than passing fads, others sustain scholarly debates over many years.

Politics also plays a role in deciding what counts as theory. For example, as Catherine Lutz (1990) argued many years ago, theory is gendered: what men write is more likely to be considered theoretically important than what women write, and thus more likely to be cited, to secure employment and visibility, and so on. We would add that authors’ structural positions play an important role: whether it deserves to or not, work by a scholar employed at a prestigious research institution invariably commands more attention than work by a scholar slaving away in multiple adjunct appointments.

Like his predecessors, AE’s current editor approaches theory in a generous and inclusive way. The main question he asks about a manuscript is, does the author draw abstractions that transcend the concrete specificities of the empirical materials? This is a key consideration for a journal like AE, which has a broad array of readers whose topical interests are diverse, who focus on different regions of the world, who work in different national contexts, and who favor a variety of methodological tools. This breadth is what gives the journal its dynamism, and maintaining this dynamism is the editor's greatest concern. Readers of AE thus cannot be assumed to have an a priori interest in the specifics of every article, although they may be prepared to become interested if the author engages them, meets them halfway. (Authors of recently submitted articles might recognize this wording from their decision letters.) And it is often this meeting halfway that constitutes a move to theory, a distillation of what is abstractly significant in a particular piece of work.


Although ethnography has been appropriated by many other disciplines and has undergone considerable critical rethinking over the decades, it is still central to sociocultural anthropology. While AE has published articles that are not based on ethnographic materials in a straightforward sense of the term (e.g., under Besnier's editorship, Gusterson 2017; Wilson 2016), the journal has a long‐standing tradition of publishing rich ethnographic descriptions that draw readers into people's lived experience. For some, this tradition may seem conservative, especially now that many anthropologists are researching field sites that do not lend themselves to traditional forms of ethnography (e.g., internet‐based communities). Moreover, conducting the kind of long‐term ethnography that has been the discipline's trademark is increasingly becoming a privilege accessible to the few. The very concept has been critically scrutinized through lengthy debates about its epistemological and ethical assumptions. Nevertheless, if anthropology is to continue to offer its unique perspective on the world and maintain its intellectual distinctiveness, it must keep nurturing ethnography as essential to the development of theory in the discipline, and AE is here to sustain that endeavor.


The last keyword is relevance, which captures our collective duty as anthropologists to shed light on the world we inhabit. Yet relevance is perhaps more politically loaded today than the other two keywords. In some national contexts, research is increasingly being evaluated in terms of “impact,” namely the extent to which research is “useful” to society and nation, with usefulness often being equated with generating income and contributing to the destructive forces of neoliberalism (Stein 2018). Our discipline often does not fare well under this kind of evaluative scrutiny, but it fares much better when relevance is understood in a much broader sense that encompasses the struggle for social justice and criticism of a status quo that favors some and marginalizes others. This is the kind of relevance that AE seeks in its articles.

Under Besnier's and his predecessors’ editorships, a steadily increasing number of articles have appeared in AE (and on the journal's website) that focus on pressing sociopolitical issues, such as refugee crises, populism, racism, and armed conflict. This is partly because we seek to foreground anthropologists’ perspectives on these urgent problems, perspectives that journalists, pundits, and scholars in other disciplines are not as well equipped to provide.

While successive AE editors have certainly not limited the journal to articles that can speak to the problems of the day, they have encouraged authors concerned with topics that are not at the cutting edge of world affairs to demonstrate how these topics can help us understand “the contemporary,” broadly conceived.

Have theory, need data; have data, need theory
The most common problem that authors have is properly establishing the relationship between two key elements of the journal's mission, ethnography and theory. Inevitably, there are many different kinds of relationships between the two. In the most basic scenario, theoretical concepts can help an author illuminate what is going on in a particular situation. For example, one might use the idea that human rights are “vernacularized” when they are transported from one national context to another, and one might demonstrate that this is happening when NGOs concerned with sexual rights operate in a particular ethnographic context. Demonstrating this has its own merit, but it does not help us understand other situations. As a result the article will appeal only to readers who are already interested in either human rights or the field site. It will thus be appropriate for a journal that specializes in the region, or one concerned with sexuality or human rights, but not for a journal like AE.

A more ambitious goal consists of applying theoretical ideas and concepts to ethnographic specificities and then demonstrating how this exercise demands that we revise our understanding of these ideas and concepts. To continue with the previous example, an article that accomplishes this would demonstrate how the specifics of the ethnographic context lead us to rethink the concept of vernacularization or the assumptions underlying sexual rights, with implications beyond the confines of the specific ethnography. A yet more ambitious goal is to make a significant contribution to theory that will create a debate and change how others think about theoretical questions. These are AE’s goals.

Two further problems often trouble authors when it comes to connecting theory and ethnography. One is that their ethnographic materials, however fascinating they may be, do not lead to a “scaling up”; that is, analyzing them contributes little to anthropological theory. This problem is particularly acute when the author assumes, in “provincial” fashion, that the ethnography is inherently interesting, especially when the field site is located in the United States or in other large‐scale societies like China. In such cases authors think they can dispense with making a case for their work's relevance and ignore its broader intellectual ramifications. Sometimes authors try to get around the scaling‐up problem by performatively asserting, “This article contributes to theory about x,” a telltale statement that immediately rings alarm bells in the editor's mind, since the author usually fails to state what the contribution actually is.

The second problem occurs just as often: authors promise high theory but provide little ethnography to support it. The manuscript ends up reading as if the author were trying on a coat five sizes too large. This kind of manuscript is often cast in overwrought prose, as if to make up for the lack of empirical evidence with obfuscating language. Much of this obfuscation comes in the form of quotations, usually from Continental philosophers and their contemporary interpreters, whether or not their ideas are relevant to the arguments.

Finding the right balance between theory and ethnography is never easy. The best approach is to think of the article as an exercise in gentle but persuasive coaxing. How, for example, does one persuade a scholar who works on the politics of national parks in East Africa to read and find something relevant to her own work on the struggles of mortgage defaulters in California? This is the litmus test for theory.

Motivating and grounding one's research

How to motivate one's work? And how to convince busy scholars with too much to read that one's article is worth reading?

Novice writers are often tempted to rely on stating that “little attention has been paid” to a topic in a particular location (“Partridge hunting in Appalachia remains an understudied phenomenon in anthropological research”). To use Edmund Leach's (1961) oft‐quoted phrase, this strategy assumes that anthropologists are in the business of “butterfly collecting” (5), that is, accumulating cultural facts in one location after the other in the manner of an encyclopedist. But just because no one has investigated a particular topic is not in itself a convincing reason to study it, since there may be very good reasons for this: studying it may require tools we do not have, or it may not produce much theoretically.

More convincing is to ground one's research in contemporary discussions and debates by referring to works in which these debates have taken place:

Distinct from—although by no means antithetical to—engaged or activist anthropology that speaks truth to power or makes its findings public (Farmer 2004; Kirsch 2014; Scheper‐Hughes 1993), this kind of public ethnography simultaneously emerges from and reflects on work with social actors who themselves work, in theory, for the public good. And, while ethnography does indeed have a public afterlife that raises new questions and debates (Fassin 2015), this public life can further participate not only in enlarging the ethnographic record but also in providing new openings for conceptual work. (Biehl 2016, 438)

But here other problems arise. Inexperienced writers (and sometimes more experienced ones) often offer “literature reviews,” reverting to a pedagogical tool that is drummed into us in graduate school. The literature review is problematic in several ways. More often than not, it is deadly boring. Because it aims to be comprehensive, its boundaries invariably exceed what the writer ends up using in analyzing the materials. It is also a sure way of badly integrating theory with ethnography, because, as a hermetically self‐contained genre, it prevents ideas from circulating from the “background” to the “situation.”

Literature reviews often take the form of “serial citations,” of which the following is an example (fictitious, but inspired by real examples):

Several scholars have explored beer consumption in the United States as a cultural site where people articulate various identities (Jones and Smith 2015), build friendships (Glover 2011), develop distinction (Bronson 2012), form consumerist subjectivities (Black 2008), and support fledgling small businesses (Baker 2015).

In this example the author says nothing about what the authors cited actually write about the topics listed (e.g., how Jones and Smith demonstrated that beer consumption articulates various identities, which is what matters). The serial‐citation structure prevents the writer from saying anything about how these various ideas relate to his or her own arguments, and whether in fact they are relevant at all. Because the passage does not provide this information, it is in fact impossible to rewrite it. More cynically, serial citations provide evidence of little more than the author's ability to search the internet.

In addition, the author wastes precious word count by stating that “several scholars have explored beer consumption,” an utterly dispensable remark. This problem is often encountered in other forms, for example, “There is a growing academic literature”; “Scholarship … has examined”; and “Anthropologists have turned their attention to x.” These statements assert only that people have written things about things.

A different strategy consists of presenting other authors’ works in serial fashion, as the following example illustrates:

Gonzalez (2015) argues that neoliberal understandings of public intellectuals and mass communication amount to participating in the “marketplace of ideas.” Bastone (2015) relates this argument to governance in universities, which are now administered as nodes within this marketplace. Grenstein (2014) more directly indicts both the neoliberal university and its articulation with the knowledge economy. And Wakefield (2011) theorizes the news media's degeneration as intimately connected to the neoliberal rollback of “the social.”

This strategy is somewhat more informative than the previous one, since it does summarize the arguments that each author developed, and it provides a sense of logical progression from one idea to the next. But it is still problematic because the grammatical subject of each sentence is an author, which creates text that is choppy, logically discontinuous, and poorly organized. The backbone of the text should not be the names of scholars (or references to “scholars,” “literatures,” “the anthropology of x,” or similar expressions), but, instead, ideas and the logical connections among them. So the passage can be rewritten as follows, which is much clearer and flows much better:

Neoliberal understandings of public intellectuals and mass communication amount to participation in the “marketplace of ideas” (Gonzalez 2015). This relates to governance in the neoliberal university, which is now administered as a node within this marketplace (Bastone 2015) and which articulates directly with the knowledge economy (Grenstein 2014). The news media's degeneration may thus be theorized as intimately connected to the neoliberal rollback of “the social” (Wakefield 2011).

Authors need to be attentive to the fact that, in English, the grammatical subject of a sentence is usually the discourse topic, so if this slot is filled with an author's name, a nominalized verb, or an abstract phrase, it will mystify agency and disrupt the continuity of topics.

A much better strategy consists of weaving other scholars’ work into the text as it becomes relevant to the arguments. And, in particular, the writer must engage with other authors’ ideas rather than simply mention them—meaning that the writer must demonstrate how these ideas are relevant to his or her arguments, how they contribute to them, whether the author agrees with them or is critical of them, and how the author's work adds to them or modifies them.

Under most circumstances, the backbone of a text should comprise ideas, as well as their relationships to one another and the relationship between other scholars’ ideas and the author's ideas. The authors of ideas or their containers (books, articles, etc.) are usually of marginal interest and thus should be backgrounded in citations at the ends of sentences. There are some exceptions, however, such as when one compares theoretical positions associated with influential authors:

For Freud, “civilization” can exist only if each individual learns to control and suppress his or her polymorphous perversions. For Foucault, in contrast, perversions are the product of society's efforts to suppress them.

In such contexts, the ideas are so closely identified with an author (or a school of thought) that it makes sense to give textual prominence to the author.

When to cite

Novice authors tend to cite far too copiously. They seem to do this out of insecurity about the validity of their own ideas, which they then try to back up with other people's ideas—erroneously believing that their arguments will be evaluated according to whom they cite. The result is what Priya Nelson (2017) calls “citational excess.”

Authors must bear in mind that citations interrupt the reading process, since they invite the reader to consult the list of references. Citing sources must therefore be done judiciously. Does the citation really add to the arguments, or is it merely a symptom of one's anxiety to show that one has done one's reading, and thus to be taken seriously?

Since our ideas always rest on other scholars’ ideas, it is essential to cite them. This will in some cases involve citing works that are canonical to certain topics. Today, for example, it is difficult to write about the “anthropology of the good” without engaging with Joel Robbins's writings on the subject or about the narrative construction of the self without some reference to Elinor Ochs's fundamental work on the topic. But there is no list of obligatory citations associated with each topic in the discipline. Experienced manuscript reviewers are very cautious about faulting a manuscript for “not having cited so‐and‐so,” and when that does happen, the AE editor (and other journal editors, we hope) takes these comments with a grain of salt, particularly when so‐and‐so is the reviewer him‐ or herself. For what specific purpose, the editor asks, should the author cite so‐and‐so's work? What ideas does this scholar present that are relevant to the manuscript author's argument? A careful reviewer will spell out why the author should engage with the missing work rather than simply state that the author should cite it. In turn, an author must make it very clear why the cited work is important to the arguments that she or he is developing. An amusing twist that occasionally arises in the double‐blind review process is when a reviewer criticizes an author under review for not having cited the author's own work; the author should take this as a sign that others see her work as important.

Since citations are attributions, to whom one attributes what idea can be a loaded issue. For example, authors sometimes attribute ideas to fashionable authors outside anthropology, even though the topic is well trodden in our discipline. This came up when an AE author attributed the idea that emotions are social to a widely cited contemporary literary theorist. Haven't anthropologists been playing with this idea since at least the 1960s? We urge authors to first turn to anthropology's rich history of theoretical debates before searching for inspiration outside the discipline.

Burying ideas: Metadiscourse and other methods of obfuscation

Article submissions often teem with first‐person statements like “I argue,” “I describe,” “I explore,” and, the worst of them all, “I look at.” Another variation removes the animate subject from the sentence and substitutes an action, yielding awkward constructions like “analyzing x demonstrates that y” or “an ethnography of x shows that y.” Such performatives (in the Austinian sense), or “metadiscursive frames,” are usually unnecessary and often counterproductive, since they license the author to tell us that he is asserting an idea without specifying what the idea is. In the following example, if we remove the metadiscursive frame (“This article explores”), we are left with only a research topic rather than the argument that readers need to know:

This article explores interactions between managers and workers at a clothing manufacturing plant in Shenzhen, China.

In other cases, metadiscursive frames are redundant and only weigh down the text. For example, compare the plodding first version of the following sentence with the much‐lighter second version:

Original: I argue that transgender people in Tonga contest the moral condemnation of which they are the object by ethnographically documenting that they claim to be part of a moral world that transcends the local.

Revised, with metadiscourse removed: Transgender people in Tonga contest being morally condemned by claiming to be part of a moral world that transcends the local. [2]

Like serial citation, such metadiscourse is often a symptom of insecurity. For example, the author might be keen to explicitly emphasize the uniqueness of her contribution, knowing that her article is fated to vie for clicks in a competitive digital marketplace. Perhaps she is reluctant to assert her arguments or has only an inkling of what they are. Or the metadiscourse is, like the literature review, a holdover from her earlier training as a graduate student or even as an undergraduate: some professors in the social sciences, in an effort to disabuse students of their “literary” pretentions, require “I argue” to force them to clarify and make explicit their arguments.

In short, authors should avoid narrating their own thought process or asserting that they are writing about something (or that another scholar has done so). They should instead state their arguments upfront and tell the story. The metadiscursive frame will then of necessity become redundant, and the connections among the author's ideas will be easier for the reader to follow.

This is not to say that one should eliminate all metadiscursive devices, since they can be crucial to structuring ideas. “The problem,” as Joseph Williams (1990) wisely counsels, “is to recognize when metadiscourse is useful and then to control it” (125). Helpful metadiscursive devices include sequencers, which help direct the reader, such as section headings and list enumerators (e.g., first, second, and third); logical connectors, such as therefore and however; and hedges, such as perhaps and seems (Williams 1990, 126–30).

Metadiscourse is not the only means by which authors bury their ideas in thickets of verbiage. Another culprit is an inflated, overly formal register replete with Latinate vocabulary. Insecurity, again, may be the cause: authors think that to be taken seriously they must use, for example, assist instead of help, utilize instead of use, interrogate instead of question, possess instead of have, or commence instead of begin. While it is true that scholarly writing demands a certain gravitas, such word choices make for stilted writing. We see this especially in manuscripts submitted by ethnographers who study bureaucracies. Having spent long hours with their interlocutors in nonprofits, state agencies, international aid groups, and the like, they produce articles cast in the dreary language of a government report or white paper—with copious doses of passive constructions, long strings of prepositional phrases, and Unnecessary Capitalization.

When terms carry particular analytical significance, authors should think very carefully about word choice. For example, verbs like point to, link to, include, and have to do with are so vague as to be essentially meaningless, yet they often appear in sentences that are crucial to the article's argument (e.g., “positions of power in this society are linked to the control of esoteric knowledge”). In addition, authors, particularly novice authors, often use terms that are fashionable (often borrowings from French and similarly exotic languages) without weighing their theoretical and stylistic implications. For example, is the term assemblage really necessary to the argument, or is it there because it sounds trendy and everyone is using it these days?

Less commonly, authors who study subalterns will, perhaps in an effort to establish their street cred or posture as “cool” fieldworkers, sprinkle their prose with slang or profanity. The best course is to chart a middle path between the dual vices of formality and informality. Use, for example, neither domicile nor digs but house. As for the language of one's field site—be it a homeless camp or a human rights bureaucracy—it will be sufficiently evoked in quotations from one's interlocutors. There is no need to adopt it as the authorial voice.

Word choice is not the only problem, however. Unnecessarily formal diction in academic writing also comprises a handful of phrasings that are so widespread that they have practically become normative. These include, for example, nominalizations, or verbs transformed into nouns: “The company's reduction of its workforce led to a rise in unemployment claims” (revised, with the verb brought out: “The company reduced its workforce, leading more people to file unemployment claims”). These constructions often result in run‐on sentences with multiply embedded subordinate clauses; reading them is akin to doing a crossword puzzle. Use the active voice whenever possible and above all avoid the agentless passive when specifying the agent is ethnographically important. [3]

Avoiding this sort of writing applies just as well to handling quotations. Many novice writers seem compelled to reverently quote scholars whose fame rests largely on their ability to render simple ideas in overwrought prose: “As so‐and‐so insightfully writes, ‘Nostalgia is the act of bringing the past into the present for a reconfiguration of the future's otherwise.’” In such cases, it is much preferable to (try to) paraphrase rather than quote the scholar in question.

We need not dwell much further on these problems, since they have been capably discussed elsewhere in writing manuals, which we advise potential contributors to consult. Particularly helpful is Williams's (1990) discussion of topicalization, nominalizations, and coherence, as well as the essay entries on composition in Bryan Garner's (2016) usage dictionary. [4] Attending to these matters, as with metadiscourse and serial citation, is not just a question of aesthetics and readability. It is also a question of laying bare the author's ideas, making it possible to clarify their interconnections and use them as the backbone of the article.

Structuring the article

A key task in writing good scholarly prose is to throw off some of the habits of one's training. This is the case with, as we've already mentioned, the literature review, serial citation, metadiscourse, and an excessively formal register. It is also the case in how we structure the article.

Many of us were taught to draft our papers according to the model known as IMRaD (introduction, method, results, and discussion) or variations on it. As the dominant model for journal articles in the STEM disciplines, IMRaD largely serves as a vessel for metadiscourse. For that reason—as well as because the IMRaD is so rigid and uncongenial to narrative—we encourage authors to avoid structuring their articles this way. Yet we cannot offer our own, alternative model, since the ethnographic and theoretical content of each article will largely determine the form it takes. We cannot overemphasize, however, that outlining the article before drafting it is an exercise much to be recommended, given how many submissions we receive that suffer from disjunctive writing (e.g., sentences that do not follow each other, misplaced paragraphs, section headings that do not describe what follows). An outline ensures that the article conveys a sense of logical progression rather than thrashing about in multiple directions. [5]

During the outlining process, anthropologists should bear in mind that they, compared to scholars in many other disciplines, have great stories to tell; storytelling is thus a great way of thinking about how to structure and present one's research. “Narrative desire is hot in all of us,” as Rachel Toor (2017) notes, and harnessing this desire is often the key to a memorable piece of ethnographic writing. [6] A light, breezy narrative style will bring an author fans; a plodding, grinding style will not. If there is humor and lightheartedness in the ethnography, use it to advantage. A storytelling approach opens up considerable freedom for the ethnographic writer, but it does have its limits. Thus ethnographic writing involves finding a balance between freedom and constraint. Novice writers can best get a feel for this by closely reading the work of more experienced writers.

Many anthropologists have capitalized on storytelling techniques since the reflexive days of the 1980s. One of the telltale signs of our discipline's “literary” turn is the now‐ubiquitous “opening hook.” Borrowed from journalism and fiction, this device usually takes the form of an anecdote or “scene setter” (close description of place or milieu) that begins the article and draws the reader in. The opening hook is a relatively new phenomenon; in researching this article, Niko Besnier conducted a (thoroughly unscientific) survey of AE articles through the decades: the opening hook is largely absent in the 1970s but becomes increasingly frequent in the 1980s. [7] This timing suggests that the opening hook may be the result of the “writing culture” critique of positivism in ethnography and the call for an experience‐near approach to writing.

Some writers, however, botch the opening hook by poorly integrating it. They start the article with an anecdotal paragraph or two, then abruptly switch to an ethnographic or theoretical discussion, leaving the reader to guess at what the connection is. The best openers do two things: they entice readers with a bit of mystery (Blundell 1988, 128), progressively drawing us in, and they metonymically evoke a central motif in what follows, leading the reader smoothly toward the article's arguments. The transition from opening narrative to the expository meat of the article should then be close to seamless. If this cannot be achieved, it is best to dispense with the opening hook and get right to the point.

In the main body of an article, one challenge is to hit the right balance of ethnography and theory, as we discussed earlier. Here, there is no golden rule. One can, however, check the manuscript to see if it fulfills three requirements. First, the theoretical problem one is addressing should be made clear in the introduction, so that the reader can determine where the arguments are heading before tackling the rest of the article. Second, the sections of the manuscript that present the ethnography must periodically refer back to the general problem, usually at the end of the section, by explaining how the ethnographic particulars presented in the section shed light on the theoretical claims. This is to avoid a manuscript structure in which the theoretical problem is the focus of the beginning and end of the manuscript, without any mention of it in between. Third, the conclusion (which must be clearly identifiable) must clearly “echo” the problem established in the introduction, so that the introduction and the conclusion do not seem to belong to two different manuscripts; this often happens when an author's thinking has evolved during the writing process but the author has failed to revise the manuscript accordingly.

The sections of the main body of the article, sandwiched between the introduction and the conclusion, should normally be about the same length, should begin with a clear introduction and conclusion, and should connect to one another with transitions. The section titles should reflect the content of the section and “speak” to one another (i.e., be similarly structured), and outlining is a great way of ensuring this. If there are compelling reasons to deviate from these norms, they should be very clear from the context.

The conclusion is also where readers expect the author to scale up arguments and provide an envoi that brings the discussion to a level of abstraction beyond the specifics of the ethnography. In fact, we suspect that, after the title and abstract, the conclusion is what readers are most likely to read before deciding whether to read the entire text (that's how the editor begins the evaluation process). A useful exercise is, therefore, to write the last two or three paragraphs so that they do not refer to the ethnography's specific location and particular topic. It is to this end that we also encourage authors not to title their conclusion “Conclusion” (what else could the last section of an article be?) but rather to give it a more descriptive title that elegantly captures the work's theoretical contribution.

Titles, abstracts, and keywords

Gone are the days when anthropologists read every article in their favorite journals. Today, print copies of journals no longer land in mailboxes, and researchers rely almost exclusively on internet searches to find articles. The centrality of search engines in contemporary research has radically changed the purpose of titles, abstracts, and keywords. Search engines use them to locate articles, and they play a prominent role in readers’ decisions about whether to read them. Likewise, scholars decide whether to serve as peer reviewers of an article based on its title and abstract. Authors must therefore take great care to write an interesting title and a readable, engaging abstract to both lure readers and help editors cajole potential reviewers. [8]

Article titles must be catchy but informative. As with the opening hook, a touch of mystery can help catch the reader's attention, and a bit of humor or a double entendre often helps. Authors can give free rein to creativity by capitalizing on the appeal of the unusual or unknown, but within limits. Article titles are not like film titles, and a wholly obscurantist title is unlikely to persuade anyone to read the article. Thus, we do not recommend using main titles that make sense only while one reads the article (e.g., “Our Neighborhood Is Not What It Used to Be”). Titles that are cryptic (e.g., “Incandescence”) must be accompanied by a disambiguating subtitle. Here are some good article titles recently published in AE: “Professional Apartheid: The Racialization of US Law Schools after the Global Economic Crisis” (Tejani 2017); “Torn Dollars and War‐Wounded Francs: Money Fetishism in the Democratic Republic of Congo” (Walker 2017); and “The Makeup of Destiny: Predestination and the Labor of Hope in a Moroccan Emigrant Town” (Elliot 2016).

As these examples illustrate, most article titles in the discipline consist of a main title and a subtitle, the latter providing specificity to the former. Here again, parsimony is of the essence: plodding, unnecessary phrases like “a case study of,” “an ethnography of,” and “reflections on” anticipate a merely descriptive article with limited ambitions.

The same requirements apply to abstracts. An abstract must summarize the main points of an article precisely and concisely, and in as readable a style as possible. We advise authors to resist the temptation of slapping together an abstract at the 11th hour, just before uploading the manuscript to the submission website. Authors also need to remember that once the article is published, the abstract, along with the title and keywords, is the only part of the article that will be freely available online if the journal is not open access.

We keep AE’s abstracts short, from 100 to 125 words long, because busy readers will not have the patience to wade through a lengthy summary, and we need room on the first page for translations of the abstract. It is essential that the abstract be crystal clear, that it be free of jargon and convoluted syntax, and that every word contribute essential information. Our warnings about unnecessary metadiscourse thus apply doubly here, which is why we prohibit beginning abstracts with “In this article.” Since readers generally read the abstract before reading the article, it must be intelligible independently of the article. It should also state precisely what the article contributes to larger questions. Thus it is inadequate to end an abstract with a statement like “this analysis opens new directions for the anthropology of x,” because this does not tell us what the new directions are.

Finally, keywords are essential to an article's online discoverability, so authors should choose them with the utmost care. Ideally, keywords are taken from the title and abstract, and they should be the object of sustained analytical attention in the main text.

Writing anthropology: Generosity and hard work

We have now addressed all the key problems that crop up time and again in manuscripts submitted to AE. Our advice has been eminently practical: get to know the journal before submitting to it, integrate theory and ethnography, avoid needless metadiscourse and inflated diction, focus on storytelling, and so on. But we have one last piece of advice to offer, a less practical one but perhaps the most important: that authors develop an ethic of generosity toward the reader. Such an ethic demands, first, that the writer imagine an ideal reader—a generous, open‐minded reader, willing to bear with the writer through several pages of perhaps difficult but enlightening material. Second, it demands that the writer empathetically adopt the position of this ideal reader, imagining what it would be like to approach the article as someone who lacks much of the author's knowledge. Once the writer commits to this ethic, it becomes second nature to think rigorously about how to clarify one's ideas and to take practical measures, such as the ones we have suggested, to ensure that they are properly presented.

Writing well is hard work, and it is perhaps especially difficult for those who are new to publishing. This may be so for several reasons. As we have noted, authors may have formed habits in graduate school that are hard to break. Or perhaps they haven't had a chance to focus on improving their writing since their undergraduate days, given that most anthropology graduate programs focus above all on producing sophisticated ethnographers and theoreticians, not prose stylists. They are provided poor models in much of the academic canon, so much of which is abstrusely written. And harried young scholars, facing a dismal job market and precarious employment, have little time to devote to studying composition.

There is little that we, as journal editors, can do to change these circumstances. The best we can say is that, despite the pressures and limits that one encounters in the neoliberal university, adopting communicative generosity as one's credo will go a long way toward producing good writing and successful articles—and, in the process, bucking some of the more unfortunate trends in contemporary academic prose.


Acknowledgments. We are grateful to Susan Brownell, Ilana Gershon, and Priya Nelson for their critical comments. We also thank the numerous authors of manuscript submissions who unwittingly encouraged us to reflect on the issues we discuss here. We would like to take this opportunity to thank the multitude of scholars who volunteer their time to anonymously review manuscripts for AE. Their invaluable expertise contributed directly to this article.
[1] For more general advice on writing well in anthropology, the following sources may also prove helpful: Becker 2007; Brown 2016; Ghodsee 2016; Narayan 2012; Van Maanen 2011.
[2] The reference to ethnography in the original version is redundant in a submission to AE because the journal's articles are assumed to be based on ethnographic fieldwork. Such a comment would be appropriate in an article submitted to a journal whose readership needs to be convinced that ethnography is a powerful tool.
[3] The passive construction is much maligned for its role in mystifying agency. But a total ban on passives is inadvisable. Joseph Williams has the measure of things when he notes that passives can play a crucial role in regulating the flow of information; see Williams 1990, 47–48.
[4] See especially the following entries in Garner 2016: “Abstractitis,” “Be‐Verbs,” “Bureaucratese,” “Formal Words,” “Legalese,” “Noun Plague,” “Officialese,” “Passive Voice,” “Prepositions,” and “Zombie Nouns.”
[5] For advice on outlining, see chs. 12 and 13 in Booth et al. 2016.
[6] Bell's (1997) guide to narrative structure, though meant for fiction writers, may prove helpful to the ethnographer.
[7] One could, however, consider this a proto‐hook from Bronislaw Malinowski's 1922) famous Argonauts of the Western Pacific: “Imagine yourself suddenly set down surrounded by all your gear, alone on a tropical beach close to a native village” (4). But this classic “arrival scene” does not occur until section 3 of the introduction, and for many different reasons we do not recommend it as a model (Pratt 1986).
[8] Cajoling reviewers is especially important. In the current climate, it has become increasingly difficult to persuade scholars, already exhausted by their teaching and administrative work (in addition to their own research), to add manuscript reviewing to their list of tasks (Grinker and Besnier 2016).