20.1 Distinguishing Audiences from Users

To say that a dictionary has “users” implies that it is a tool, like a calculator or a map, whereas to say it has an “audience” suggests that it is a piece of media content, like a film or novel. Both frames can offer insights for lexicography, but the concept of an audience is particularly helpful when considering unspecialized monolingual dictionaries, whose perceived value is grounded in ineffable social and cultural qualities as well as concrete utilitarian ones.

In general, the more specialized a dictionary is in meeting the needs of a particular type of user performing a particular task, the more susceptible it is to being evaluated as a tool. Tools can be evaluated based on their efficacy in enabling specific activities, and dictionary user studies have provided powerful insights into how dictionaries can better support the needs of users, especially language learners and translation professionals. Monolingual dictionaries give representative coverage of the vocabulary for general consultation and provide a range of information not only about words themselves but also pertaining to general knowledge. For non-professional users who casually consult them, monolingual dictionaries can be seen as functioning as a form of mass media, particularly when accessed as freely available online resources. When used in this way, the quality of a dictionary cannot be assessed in terms of efficacy alone but has a more complex relationship with audience expectations and social context.

A focus on dictionary audiences might examine the complexities of the relationship between dictionary producers, dictionary content, and dictionary users within broader social, political, and economic contexts. The interplay between dictionary makers and audiences has shaped the dictionary over time, as marketing and business decisions both create and react to consumer demand. Consciousness of the global variability of “dictionary cultures” is also a crucial underpinning for consideration of the dictionary audience. M. Lynne Reference MurphyMurphy (2018, 2) notes that “dictionary cultures reflect and contribute to the larger society’s situation concerning language, authority, literacy, education, commerce, and interactions of these and related areas, and the complexes of social values concerning these matters.” Intentionally or not, dictionaries carry messages about much more than just words. At a time when the established twentieth-century dictionary landscape has been thoroughly destabilized by developments such as the decline of print, the rise of free internet resources, the advent of new licensing business models, and increasing hybridization and fragmentation of dictionary texts, dictionary publishers must reconsider their relationship to dictionary audiences as a matter of survival. The field of lexicography studies has potential to shape the future of dictionaries by developing theories and methodologies to support such reassessment.

20.2 Historical Audiences

Histories of lexicography emphasize the development of approaches and methodologies over time and the contributions made by different works and individuals more than their reception by a mass public, but an implied audience is often present at the backdrop of those narratives. Lexicography, broadly understood, is as old as writing itself, but while ancient wordlists, glossaries, and dictionaries certainly had users, such as scribes, poets, clerks, students, or clerics, there can’t truly be said to be an audience for dictionaries until they became available for mass consumption as published books.

In 1760, Benjamin Franklin wrote to David Hume that “it often gives me pleasure to reflect how greatly the audience (if I may so term it) of a good English writer will, in another century or two, be increased.” This self-consciously novel usage of the word audience supplies the OED’s first citation for the word as used to refer collectively to the readers of a particular publication or writer, extending the word’s application to the listeners or spectators who were physically present at a live event to denote an asynchronous mass of readers. Widespread literacy and commercial publishing created demand for this meaning of audience, just as they created demand for dictionaries.

Printed dictionaries, like any printed book, require investment to produce and circulate, necessitating that their producers have in mind some audience or market. Whether the business model is patronage, subscription, or direct sales, the audience must be defined and appealed to, and those appeals provide insight into how dictionary makers define their public (see Morse, Chapter 29, this volume, for detailed discussion of the evolution of commercial dictionary business models in the United States). In this way, the intellectual history of dictionary makers and the commercial history of dictionary publishers contain within them insights into the history of audiences. The general trajectory of that narrative of audience development (who was using dictionaries, and why) is well known for dictionaries in the United Kingdom, the United States, and some Western European countries, but further illumination of its contours in countries and languages with very different lexicographical traditions would be valuable for understanding how global audience expectations may differ today. Reference ConsidineConsidine (2019d), the first global history of lexicography, reveals the immensity of the topic, recounting the distinctive developments of more than a dozen different modern lexicographical traditions. With that diversity in view, the impossibility of any overarching narrative about dictionary audiences is obvious, but also obvious is the potential for rich comparative work to understand the peculiarities as well as the commonalities in dictionary cultures. It is far beyond the scope of this chapter to attempt to address that gap, but some examples touching on how monolingual dictionary producers perceived and cultivated audiences are briefly reviewed.

Early dictionary titles and prefaces often contain references to potential users and, consequently, reward critical reading; Reference ShapiroShapiro (2017) collects explanatory frontmatter from early English dictionaries. The work generally regarded as the first English monolingual dictionary, Robert Reference CawdreyCawdrey’s 1604 Table Alphabeticall, is a collection of 2,500 “hard usuall English words” defined using “plaine English words.” Its verbose full title explicitly invokes a non-professional audience of “Ladies, Gentlewomen, or any other unskilfull persons” who are perceived to have a need to understand difficult words they hear or read “in Scriptures, Sermons, or elsewhere” and may also wish to use themselves. The image is of private, domestic use of the reference, as opposed to the formal didactic or professional use associated with bilingual dictionaries.

Much has been made of the specific mention of women here and in the title pages and dedications of other English lexical reference works during the seventeenth century, with the suggestion that women were conceived of as a key part of the early dictionary audience. Reference RussellRussell (2018, 28–67) questions whether the well documented invocations of women in frontmatter represent genuine demand from an uneducated female audience in need of educational tools, as opposed to an appeal to them as influencers to promote and endorse a novel product, the dictionary. Her analysis of how women are invoked in fourteen dictionaries cited by James Reference MurrayMurray (1900) in his Romanes Lecture shows a clear distinction between multilingual and monolingual dictionaries in this respect. The title pages and dedications of multilingual dictionaries overwhelmingly cited specific prominent female individuals, typically with relevant linguistic expertise. In contrast, monolingual dictionaries were much more likely to specify women, undifferentiated, as a class of potential readers – an audience. By the eighteenth century, these specific invocations of women disappeared, as monolingual dictionary consumption entered the sphere of scholarship, literature, and criticism.

During the seventeenth and eighteenth centuries the audience of the monolingual dictionary expanded to include scholars and educated literary men, as vernacular monolingual lexicography developed extensively in various Western European countries, with dictionaries playing a role of celebrating, cataloging, and standardizing national languages. Language academies sponsored large and ambitious dictionaries codifying literary vocabulary in Italy (1612), France (1694), and Spain (1726–1739). While England had no academy, Samuel Johnson’s A Dictionary of the English Language (Reference Johnson1755) sought to fill a similar function for the English language. With the elevation of monolingual lexicography to a topic of literary and scholarly concern, an audience of educated consumers and critics, with access to means of publication in their own right, has left ample record of their views.

Reference Considine and OgilvieConsidine (2020) points out that while Johnson’s dictionary was an epochal development in English lexicography, from the perspective of the English dictionary market it was only filling one particularly high-profile and elite niche in a diverse “dictionary ecosystem” dominated by cheap lexical resources targeting specific types of users, such as spelling books for children and cheap, derivative texts which were not influential in national literary and cultural discourse. In the nineteenth century, lexicographers elsewhere like Noah Webster in the US and Pierre Larousse in France introduced a new category of monolingual dictionary that aimed to be both methodologically rigorous and popular, educating middle class citizenry as a domestic object suitable for any home; in many countries, these monolingual dictionaries for general users displaced other types to become “the dictionary, the prototype of the genre” (Reference BéjointBéjoint 2016, 8). Sales figures in the millions of copies indicate the success of dictionaries pitched at general users in satisfying a need for their audience of consumers.

To be the dominant monolingual dictionary in a market, the one that people think of when they talk about the dictionary, is an enviable position, and one worth fighting for from a commercial perspective. The term “dictionary war” has been applied to periods of particularly intense conflict between publishing rivals for the reputational and financial benefits of being seen by the public as the best or default dictionary choice. The foundational such conflict in American lexicography took place in the mid-nineteenth century, when Webster’s lexicographical legacy was challenged by Joseph E. Worcester, whose dictionary was more faithful to British norms, in a decades-long acrimonious debate in the press (Reference MartinMartin 2019). Another round took place in the 1960s, when The American Heritage Dictionary of the English Language (Reference Morris1969) embraced a role as a guide to accepted usage, positioning itself as a responsible alternative to the controversially permissive Webster’s Third New International Dictionary of the English Language (Reference Gove1961) (Reference 697AdamsAdams 2015). Reference 714DollingerDollinger (2019) has called the competition between Gage Publishing and Oxford University Press in the 1990s for primacy as the preferred dictionary of Canadian English “The Great Canadian Dictionary War.” In all of these cases, dictionary publishers leveraged broader social, cultural, and political dynamics to make the case for the superiority of their product; the debate over dictionaries resonated with the public because it was ultimately about much more than just dictionaries.

M. Lynne Reference MurphyMurphy (2018) sees links between American characteristics like belief in social mobility and reverence for textual authorities like the United States Constitution, on the one hand, and key messages promoted by Merriam-Webster marketing materials in the nineteenth and twentieth centuries, on the other. That aspirational view of the dictionary’s role is contrasted with British dictionary advertising by Oxford University Press, which seems designed to appeal to audiences that were already part of an elite. It is natural to wonder whether the images of the dictionary promoted by marketing departments also shape the content produced by the editorial departments.

Commercial competition among publishers with its attendant marketing and public relations campaigns both affects and is affected by audience perception of what the role of the dictionary should be. Researchers in the communications field have questioned the traditional binary division between production and reception of media content, seeing the audience as having agency in shaping what is produced in order to appeal to them (Reference RossRoss 2014). This line of thinking suggests that, instead of one-way communication between the dictionary publisher (and their marketing department) and the dictionary audience, we should imagine a feedback loop in which the public also wields active influence.

20.3 Current Audiences

To understand contemporary dictionary audiences, it is natural to begin with insights from research about dictionary users. The goal of dictionary user research has been summarized as “to study how human users interact with dictionaries with the aim of making this interaction more effective (improving success), more efficient (faster), and more satisfying (pleasant to use)” (Reference Lew and Gilles-MauriceLew and de Schryver 2014, 344). User-oriented research has its earliest roots in the 1960s but began to coalesce as a discipline in the 1980s, to the extent that in his review of the aptly named collection Lexicography: An Emerging International Profession (Reference IlsonIlson 1986), John Reference AlgeoAlgeo (1986, 262) described the progress in recent years as amounting to “a quantum leap.” The flourishing of dictionary user studies starting in the 1980s coincided with several other exciting developments in the field of lexicography – the adoption of corpus-based methodologies, the availability of new digital methods for dictionary compilation, innovations in the field of monolingual dictionaries for learners, and the advent of electronic dictionaries – providing ample fodder for decades of research (for recent overviews of the field and its methodologies, see Reference Nesi and JacksonNesi 2013; Reference SvensénSvensén 2009, 453–479; and Reference TarpTarp 2009).

Reference Varantola and CorréardKrista Varantola (2002) offers a simple typology of dictionary users: language learners, professionals, and non-professionals. User research has tended to focus on the first two categories, in particular on the functional needs of university students studying English, for whom effectiveness of dictionary consultation can be evaluated through small-scale experimental and observational studies considering quantifiable outcomes such as speed of finding information, accuracy in completing a task with a testable result, or success of learning new information. Because of this focus, use of bilingual and learners’ dictionaries has been investigated more thoroughly than use of general monolingual dictionaries, and when monolingual dictionaries are considered, it has often been with consideration to their utility for L2 users, not L1 native speakers.

Research into monolingual dictionary users has often been conducted using questionnaires. The questionnaire method has many shortcomings, and comparisons are perilous when made between different studies conducted with different samples, even when questions are superficially similar. Nonetheless, as Reference LewRobert Lew (2002, 268) notes, questionnaires can enable research into aspects of dictionary use that are difficult to investigate by other means, such as historical and attitudinal aspects. When considering audience attitudes and dictionary cultures, the problem of unreliability in self-reporting is mitigated by the fact that how individuals recall and represent their use of dictionaries is just as relevant as how they actually do use them. The questionnaire method also has the advantage of being scalable to large numbers of people, which makes it helpful for considering perspectives within a population.

Analysis of log files of digital dictionaries is another technique suited to understanding dictionary users as an audience, since it offers a perspective on natural usage that is not shaped by the observer and can potentially provide an overview of millions of users if the dictionary is sufficiently popular (although access to such information is often proprietary in nature, which inhibits its publication). Social media can also be used as a source of information about what different segments of the public collectively want from dictionaries; Reference Bae and NesiSusanna Bae and Hilary Nesi (2014) compare questions posted by Korean and English users on online Q&A sites. This approach, like the use of log files, enables observation of real-world behavior, and it also has the advantage of providing a window onto the attitudes of people who may not actually be using a dictionary yet, for example those who are trying to select one or who aren’t sure if a dictionary will address their needs.

A survey of American teachers at ninety-nine colleges conducted in 1955 by the dictionary editor Clarence Barnhart is considered the first such study of dictionary use. The respondents were asked to rate six types of common dictionary information according to their importance to college freshmen; the results rated meaning first, followed closely by spelling, then pronunciation, synonyms, usage notes, and, in a distant sixth place, etymologies (Reference Barnhart and AdamsBarnhart 2022 [Reference Barnhart and Adams1962], 279–280). Studies of English-language monolingual dictionary users consistently find that looking up meanings and spellings are the most common reasons to consult a monolingual dictionary, and that using dictionaries to settle disputes and as a reference for word games are also important roles (Reference Greenbaum, Meyer and TaylorGreenbaum et al. 1984; Reference SiegelSiegel 2007; Reference SvensénSvensén 2009, 465). The idea of using a dictionary to “settle a dispute” is one that comes up consistently in discussion of monolingual usage. A student respondent in Muffy Siegel’s study explained “I usually go to a dictionary if my family is having a debate over the proper use of a word, such as the difference between insure and ensure” (Reference Siegel2007, 37).

Conducted in 2017, The European Survey of Dictionary Use and Culture (Reference KosemKosem et al. 2019) is the largest study of monolingual dictionary use to date, involving nearly 10,000 respondents in twenty-nine countries. Its results broadly correspond with earlier findings, even though it involves respondents from a variety of ages and backgrounds and disparate dictionary cultures. The most common situation in which respondents said they used a monolingual dictionary was to look up an unfamiliar word, followed by “resolve a dispute,” and various writing and editing tasks. Playing word games was the least common use, but 17 percent selected that option. Overall, respondents identified reliable content, up-to-date content, and being easy to use as the three most important qualities for a monolingual dictionary in their language, but there were also some interesting variations in specific countries; for example, Finland, Italy, and Macedonia ranked linking to corpus data among the top three features.

The survey also probed questions about attitudes toward dictionaries, which reveal some interesting disparities between countries. Asked to identify characteristics that people associate with the leading monolingual dictionaries in their country, 63 percent of Dutch respondents selected “authority,” but only 12 percent of Macedonian respondents. The characteristic “something mostly academics care about” was selected by 47 percent of Czechs but only 2 percent of respondents from the Basque Country. The major monolingual dictionary was seen to be a national symbol by nearly 30 percent of Serbs, Georgians, and Slovenians, but fewer than 7 percent of respondents from Portugal, Spain, and Belgium. The data collected in the survey, along with extensive supplementary reviews of the situation for monolingual dictionaries in each country, is an exemplary source for considering dictionary cultures from a comparative perspective. Fault lines are exposed between, for example, the countries where orthographic dictionaries dominate and those where general monolingual dictionaries represent the language norm, or between those in which publicly funded institutions are the major publishers and those where commercial publishers dominate. One can only hope that similarly ambitious work might be carried out on other continents to further understanding of the variability of dictionary cultures globally.

Not all dictionary use is driven by the need to retrieve specific lexical information. Some users browse monolingual dictionaries simply for enjoyment. Despite being dense in information and difficult to use, prestigious academic dictionaries often have a non-professional audience as well as a scholarly one. Reference Considine and FontenelleConsidine (1998) identifies a range of ways in which historical dictionaries, specifically the OED, provide “pleasure” to their owners (see also Stavans, Chapter 24, this volume). Recreational readers of academic dictionaries rejoice in the wide range of their contents, which as part of their lexical coverage often provide encyclopedic information on a vast number of areas of human history and endeavor, and thereby seem to offer a wide-ranging education.

Readers also express aesthetic and literary appreciation for the illustrative quotations. They report browsing and reading the dictionary as a form of self-education or entertainment, and some actively engage in a community of contributors, sending notes to the editors with potential improvements. For owners of dictionaries, there may also be an element of prestige and pride at work. Large scholarly dictionaries are expensive, and to display one in one’s home suggests an investment of both money and time in erudite pursuits. In April 2020, when many celebrities conducted interviews from their homes due to the coronavirus pandemic, the actors Tom Hanks and Cate Blanchett were each spotted with twenty-volume editions of the OED prominently positioned behind them.

If we accept that dictionaries fulfill audience needs that go beyond their utilitarian functions and draw on their cultural and social resonance, what does that mean for their primary function of explaining the meaning of words? The prototypical example of a definition lookup scenario is when a reader encounters an unfamiliar word, such as a technical term, an uncommon or archaic word, or a neologism; words that cause confusion or are associated with usage problems are another salient category (Reference SvensénSvensén 2009, 467). Dictionary usage data tends to confirm insights from user research, but it reveals additional nuance.

Reference SokolowskiSokolowski (2014) presents an analysis of usage log data from the Merriam-Webster Online Dictionary, one of the world’s most prominent monolingual dictionary websites. There is a consistent set of English words that are perennially popular, and these lookups correspond to what we would expect based on dictionary user research. Among the very highest-traffic entries are affect and effect; this distinction is a notoriously difficult one and likely relates to a text production context in which the users are checking usage. Many of the entries that are most consistently in the top lookups likely fit the category of checking the meaning of an unfamiliar word, either because of encountering it when reading or, Sokolowski suggests, as a vocabulary-building exercise. These are formal adjectives derived from Greek and Latin roots, like pragmatic, holistic, ubiquitous, and esoteric, abstract nouns whose meanings are challenging to pin down, like integrity, empathy, and diversity, and fundamental but disputed political concepts like socialism, democracy, and fascism. Sokolowski also sees evidence of school writing assignments in the semiannual spikes in traffic to the entry for culture that occur at the beginning of each academic term, alongside other school-related terms. The Merriam-Webster log files even register the often-mentioned use of dictionaries to play word games: each year on Christmas afternoon, the entries for qi and za suddenly rise in traffic, as families argue over whether they are playable in Scrabble.

Topical events also noticeably impact the words people look up. Some entries that spike in traffic in the wake of an event fit the familiar category of rare unfamiliar words, like cortege after Princess Diana’s funeral, temblor after an earthquake, or triage after 9/11. But often relatively familiar words pop up in the wake of current events. Sokolowski contrasts marriage, which is looked up frequently after relevant political campaign events or court decisions, and abortion (an equally salient topic in US politics), which is not. The reason for the difference isn’t that Americans don’t know what concept the word marriage denotes; it is because the precise definition of that concept has been under dispute as marriage rights extend to same-sex couples (see also Finegan, Chapter 19, this volume, and Adams, Chapter 30, this volume). In any jurisdiction, it is the statutory definition of marriage in law that determines who can be joined in such a union, but people are still consulting the dictionary as an authority. Just as when traffic to the entry for love spikes each Valentine’s Day, it is clear that the need users are seeking to gratify by means of dictionary consultation is about something other than a word’s denotation; it involves their relationship to their society, politics, and emotions. And the question of how dictionaries fulfil those needs, or whether they even should, is not easily addressed by the current frameworks of lexicography.

In cultures that regard the monolingual dictionary as an important authority on meaning, choices made in definition-writing can have a real-world impact that is far from academic. Commercial dictionary definitions are increasingly cited in US court decisions (Reference KimbleKimble 2021). Decisions by dictionaries about questions like whether to include nonbinary pronouns, or to reject use of the generic masculine, can be experienced by audiences as either affirming or challenging their identity and values. For decades dictionaries have periodically been targeted by organized campaigns that seek changes to entries they regard as offensive or inappropriate (Reference MartinMartin 2021; Reference RussellRussell 2021); whether such campaigns target the dictionary solely because of the perceived impact of the content itself, or with a degree of opportunism because of the press attention such efforts often enjoy, the motivation is the perception that the monolingual dictionary as an institution possesses power and influence.

It is uncomfortable for lexicographers to consider how their work will impact its audience socially or emotionally; modern lexicography describes lexical behavior based on objective analysis of evidence, and the idea that the lexicographer should have any other considerations in mind sounds troublingly subjective. But it is possible to create a factual entry that is also sensitive, cognizant of how audiences have reacted to other entries in the past, and of the debates taking place in society. Kory Reference StamperStamper (2021, 93) observes that there is a gap between how lexicographers regard their definitions and how audiences, who often do not have a high level of dictionary literacy, receive them, creating potential for hurtful impact from “a single-statement equivalency definition, as in ‘black: wicked, evil or impure’” without further context. She sees a responsibility for general monolingual dictionaries to be more expansive in the information they provide and to give more attention to the connotative and relational aspects of language.

20.4 Audiences and the Future of Dictionaries

User research has tended to find benefits associated with greater specialization of resources for particular types of users and activities. For example, directional bilingual dictionaries optimized for the production needs of only one of the languages are more effective than traditional bilingual, bidirectional dictionaries intended for speakers of either language (see Reference Haas and AdamsHaas 2022 and Reference Harrell and AdamsHarrell 2022 for the origins of the directional argument), and different types of example sentences are required for reception than for production (Reference Lew, Hanks and Gilles-MauriceLew 2015a). The natural progression of greater specialization is toward complete customization of the user experience to the needs and specific situation of the individual, combining multiple sources and types of lexical data and presenting them in the optimum form for the user’s present need. Such a dictionary, divorcing the user’s experience from the underlying data and creating an output that exists in a given form only at the point of access of an individual user, has been imagined since the 1990s (see Reference de Schryverde Schryver 2003, 163 and 189). The technological reality has now progressed to the point that it undermines the very notion of a singular “dictionary.”

Current trends in the dictionary landscape suggest that the impetus to provide users with the ideal individualized solution for their specific scenario of use could be simultaneously driving both consolidation (from the perspective of dictionary producers) and fragmentation (from the perspective of end users). The print paradigm rewarded niche publishing as a way of microtargeting different market segments at different price points, with size of the physical book being a key consideration driving its content, and its method of organization being another. But traditional inclusion decisions about how to balance the number of dictionary entries (breadth) and the amount of information given in each (depth) are not relevant for digital formats. Similarly, distinctions based on macrostructural organization such as where to place multiword expressions, whether to introduce topical sections, or even traditional genre distinctions between the general dictionary and the subject dictionary, or the dictionary and the thesaurus, are rendered irrelevant when data is accessible by search and intermixed and hybridized into new forms by application creators who license the content. In this environment, dictionary producers are incentivized to focus, not on disparate dictionary datasets, but on a holistic approach to creating and maintaining lexical content that supports a wide range of computational applications which often will not be delivered by the publishers themselves. Judy Reference PearsallPearsall (2013, 3) describes good content strategy for publishers in this situation as demanding “a relatively stable core content hub that is flexible, reusable, connected, sustainable, and efficient.” Meanwhile, a dictionary user may encounter a singular source of underlying content refracted into a multitude of different product experiences without realizing it is the same dictionary, or even that it is a dictionary at all.

Not only may a singular dictionary be available for lookup in a variety of apps, handheld devices, search engines, and websites, but its content may also support features in applications that are not primarily reference works. And applications that incorporate lexical data will in many cases solve users’ traditional dictionary problems more efficiently than a traditional lookup interaction. For language reception, a simple example of this is the provision of dictionary definitions in an e-reader or e-learning platform. This feature enables a reader to easily highlight a word and receive a definition or translation from a default dictionary that is built into the application. For language production, word processors, grammar checkers, and writing assistants draw on various sources of data to automatically recommend changes to writers, who now need not actively consider whether they should consult a reference for confirmation of a spelling or point of usage. The ready availability of free translation engines resolves many L2 speakers’ questions without their needing to consult a formal dictionary. Players of digital word games need not resort to a dictionary to determine the validity of a play; the list of accepted words is built into the application. Search engines analyze queries and automatically present dictionary or thesaurus data to users based on behavior of millions of other past users, identifying that an individual’s specific need can be addressed with dictionary or thesaurus content even when the searcher does not deliberately seek it out.

Meanwhile, for complex, detailed, and nuanced questions, the Internet can easily lead to a blog post or teaching materials rather than a dictionary entry, and for the users, this is often beneficial: if I’m interested in the origin of a word, then a lively 1,000-word column in an online magazine may fit my needs better than a cryptic dictionary etymology, and if I’m confused about the difference between affect and effect in English, then a conversational blog post offering examples and mnemonics is probably a more effective learning resource than a formal usage note at the foot of a dictionary entry. But in an online environment where lexical information can be found everywhere, the problem of lack of dictionary skills as a barrier to effective consultation is replaced by a new challenge: lack of critical media skills to identify which resources are trustworthy.

Henri Béjoint sees the decline of print as a harbinger of the ultimate demise of the monolingual dictionary for general users, arguing that online sources “will provide the same information, and more, but will not play the same social role” (Reference Béjoint2016, 23). But if the social role of the monolingual dictionary is still valued by audiences, then there remains some hope that lexicography can adapt so as to continue to fill it; certainly, the dictionary producers that remain active today are striving to do that. Monolingual dictionaries have responded to the current moment by expanding their editorial footprint beyond traditional lexical content into branded blog posts, educational or entertaining ancillary materials, podcasts, videos, and social media accounts connected to their overall brand. Public relations efforts like declaring a Word of the Year are another way of maintaining the dictionary’s relevance to a public audience. This is a constant negotiation, and a new burden on dictionary editors, who are much more publicly visible and accessible and who face potential media escalations at any time – not just in the flurry of a new print edition’s press release, as in decades past.

Today’s global reassessment of categories like race, gender, sexuality, and nationality places dictionaries in a critical position, as they are targeted as a symbol of authority by groups with different perspectives on these key questions of our time. Reference BéjointBéjoint (2016, 20) characterizes definitions in a general-use monolingual dictionary as being “ideologically as close to the dominant values of the society as possible, to accommodate the largest number of users.” The impersonal, clinical, scientific perspective adopted may be “boring,” but it also enables the dictionary to maintain its status as an objective authority. In Reference MartinMartin (2021), I identify over two dozen dictionary-related media controversies and campaigns in a recent two-year period originating in both progressive and reactionary responses to dictionary treatment of gender and race, and concerning German, Spanish, Italian, Portuguese, and English dictionaries, almost all relating to digital outputs, not print ones. It isn’t comfortable for dictionary makers to be in the spotlight of public controversy, but if the dictionary ceases to be seen as a salient target in culture wars, that may be a sign that it has lost its relevance.

Specialist dictionaries need not reach a mass audience in order to have an impact on their field, but the general monolingual dictionary is by its nature a mass product, not a niche one, and relies on the goodwill of the public in order to be maintained as a resource. Professional lexicography is resource-intensive in terms of both human and technical resources, and dictionaries of current language are constantly fighting obsolescence as they chase the inexorable progress of language change. Whether invested in by commercial publishers to earn profit, underwritten by big technology companies to meet demands from their end users, or funded by public institutions as a service to citizens, the perception of public value to an audience is ultimately the driver of any major lexicography initiative.

The future of the dictionary lies not only in meeting the needs of users, but also in meeting the expectations of audiences. That is a trickier proposition that requires a lexicography attuned not just to factual evidence but also to social context. Practical lexicographers and dictionary publishers are intrinsically inclined to regard the users of dictionaries as an audience, because the success of their professional endeavors is evaluated through audience-oriented measures such as sales, traffic, and social media and press reactions. Sales and traffic figures are typically not accessible to researchers because of their commercially sensitive nature, but social media can enable access to large and diverse populations for audience-oriented research, and the discussion of dictionaries in popular media is also an important barometer of public perception. Academic lexicography studies can play a role in ensuring the survival and integrity of the lexicographical enterprise by deepening our understanding of the role dictionary users play as members of an audience.

21.1 Introduction

“To recover what people in the past meant by the things they said and what these things ‘meant’ to them is the object of intellectual history,” according to John Burrow, the first Professor of Intellectual History at the first English university to offer degrees in the subject.Footnote ¹ That generous conception of intellectual history immediately suggests its reciprocal relevance to lexicography.

Reciprocal, and multiplex. “Intellectual history” can refer to so many different (though ultimately related) activities, with which dictionaries may be concerned in so many different (and not necessarily related) ways, that an account of dictionaries and intellectual history must refrain from riding, like the man in Stephen Leacock’s story, madly off in all directions.

One path this chapter takes is indicated by a signpost which reads Noli altum sapere “thou hast no reason for pride.”Footnote ² That was one of the mottoes of an intellectual institution, the sixteenth-century publishing house of the Estienne dynasty. The history of intellectual institutions and communities is one kind of intellectual history, a kind in which dictionaries are certainly involved. But before following that path, we shall look at the signposts which indicate three others – they read “Archaeologus,” “The hedgehog and the fox,” and “Begriffsgeschichte” – and after following it a little way, we shall look toward some wider horizons.

21.2 Archaeologus: The Contents of Dictionaries as an Object of Study

Archaeologus is the title of an encyclopedic dictionary of the medieval Latin vocabulary of law and the Church in England, compiled by the antiquary Sir Henry Reference SpelmanSpelman (1626). It included a long entry for the important word feodum or feudum, in which Spelman set out his understanding, driven by examination of the Latin word in the original sources, of the origins and development of English feudalism. “The feudal relationship as thus defined […] could now be employed to bring about a radical reinterpretation of the whole body of English law as it had existed in the Middle Ages […]. This was the beginning of the genuinely historical study of English institutions” (Reference PocockPocock 1957 / 1987, 102). Here, one of the entries in a particular dictionary makes it a key text in the argument of a classic of intellectual history; indeed, J. G. A. Pocock’s use of Spelman’s entry feodum in his classic The Ancient Constitution and the Feudal Law is singled out for attention in a recent overview of the whole field (Reference WhatmoreWhatmore 2016, 38).

A dictionary made on the other side of the Atlantic, in the century before Spelman’s, Alonso de Molina’s Spanish–Nahuatl Vocabulario en lengua castellana y mexicana (1555; second edition 1571), has been seen as a source for the thought of sixteenth-century speakers of Nahuatl, one of the documents “which allow us to reconstruct a conceptual map of Nahua visuality” (Reference 721HamannHamann 2015, 86). Molina, adapting the Spanish–Latin dictionary of Antonio de Nebrija, in which there was a single entry relumbrar, o reluzir ‘to shine or glitter,’ knew as a bilingual speaker of Spanish and Nahuatl that the Nahuatl words for human perceptions of reflected light were more complex than this. So, in the first edition of his dictionary (1555, fo. 212r), he constructed the following lovely Spanish headphrase as an equivalent for Nahuatl cuecueyoca:

Further Spanish–Nahuatl entries in the 1571 edition expanded Molina’s coverage of the semantic field of reflected light in Nahuatl and were joined by counterparts in an added Nahuatl–Spanish section (Reference MolinaMolina 1571, part 1, fo. 103v, and part 2, fo. 25v). Such entries “preserve indigenous ways of understanding the world not even recorded by indigenous systems of communication” (Reference 721HamannHamann 2015, 86): they recover what early speakers of Nahuatl meant by the things they said, and how these things had meaning for them. The discussion of Molina’s dictionary by Byron Ellsworth Hamann from which my example is adapted (see Reference 721HamannHamann 2015, 98) is surely the work of an intellectual historian. Likewise, Molly H. Bassett’s account of Molina’s treatment of a set of forms including Nahuatl teixiptla – he glosses this word as ‘image[n] de alguno, sustituto, o delegado’ or ‘image of someone, substitute, or delegate’ (Reference MolinaMolina 1571, part 2, fo. 95v; discussed by Reference BassettBassett 2015, 53) – which as part of an argument of hers about the Nahua understanding, past and present, of deity embodiment, belongs to the realm of Aztec intellectual history.

Other examples can readily be adduced. For instance, the statement that “Johnson’s writings in The Literary Magazine point in the same direction to the lexicon of political terms embedded in his two folio volumes of 1755” (Reference Clark, Clark and Erskine-HillClark 2002, 120) uses entries in Samuel Johnson’s Dictionary as part of the evidence for an argument about the development of Johnson’s thought. Moreover, the rich paratexts of Johnson’s Dictionary, particularly its preface, are also relevant to the history of eighteenth-century Anglophone thought. Indeed, the preface has a special place in the history of the discipline of intellectual history, as the text in which the name of the discipline is first attested, though not quite in its current sense:

In other studies, the content of dictionaries has sometimes been seen explicitly as a subject for intellectual history, as in Audrey Truschke’s “Defining the other: An intellectual history of Sanskrit lexicons and grammars of Persian” (Reference Truschke2012). For Truschke, early Indian accounts of the Persian language, including dictionaries, are “one of the prevailing modes through which Indian intellectuals repeatedly addressed the expansion of Persian language and culture on the subcontinent” (Reference Truschke2012, 637) and “demonstrate that Sanskrit authors over the course of several centuries thought deeply and diversely about the implications of Persian for their intellectual tradition” (Reference Truschke2012, 661).

A contrasting case is Walter Hakala’s Negotiating Languages: Urdu, Hindi, and the Definition of Modern South Asia. In this book, work which would, like Hamann’s and Bassett’s on Mesoamerica, be recognized as intellectual history by many readers – and has obvious affinities with Truschke’s article – has not been labeled as such by its author. But even without the label, Hakala places dictionaries in intellectual history, and indeed in that vigorous subset of intellectual history which is the history of political ideas, when he writes (Reference Hakala2016, 32) that

The concluding methodological remarks in Negotiating Languages address two ways of missing the full potential of the intellectual history of dictionaries. On the one hand, Hakala notes the practice of “mining lexicographic materials for insights into […] material and administrative history” by historians who have “devoted little attention to the formal features of these works” (193). On the other hand, he remarks that “works devoted exclusively to the reconstruction of a history of lexicography […] have hesitated to connect the objects of their studies to the social and cultural formations that produced and were themselves documented by these works” (193), the gentlest of hints at how dry and boring (my words, not his) the history of lexicography may become when it is divorced from intellectual history.

There are times when historians of lexicography might do well to look up from the dictionary page and see the green hills of intellectual history just out of the window. Conversely, there are times when intellectual historians might do well to spend more time reading dictionaries. It is always easy to pounce on apparent omissions from even the most extensive reference works. But perhaps it is not unfair to observe that although large etymological dictionaries were important features of the Byzantine lexicographical landscape (see Reference ValenteValente 2019, 255–260); and although the ancient role of etymology as a “tool for thinking,” which has been so brilliantly expounded by Ineke Reference Sluiter, Montanari, Matthaios and RengakosSluiter (2015), was still active in those dictionaries; and although the vision of the past as standing in a living and illuminating, even authoritative, relationship with the present which is suggested by a taste for etymological dictionaries might well have interesting echoes elsewhere in Byzantine intellectual life – nevertheless, the Byzantine etymologika appear not even to be mentioned in the generally admirable Cambridge Intellectual History of Byzantium (Reference Kaldellis and SiniossogluKaldellis and Siniossoglu 2017).

21.3 “The Hedgehog and the Fox”: The Form of Dictionaries as an Object of Study

So, the content of dictionaries makes them objects of interest to the intellectual historian. And, as Hakala remarks, their form does as well. Stefan Reference ColliniCollini (2016, 17) has expressed the hope that “a wider range of intellectual historians might become more attentive to, and properly value […] the ‘literary’ or ‘formal’ properties of the texts they discuss.” Let us consider an example.

Isaiah Berlin was not simply an intellectual historian, but his essay “Lev Tolstoy’s historical scepticism” (Reference BerlinBerlin 1951) is a masterpiece of intellectual history. It is better known by the catchier title under which it has been reprinted, “The hedgehog and the fox.” The neatness of the fox / hedgehog dichotomy which introduces the essay, and the fun of applying it to given persons, has made the dichotomy more famous than the central argument, which is about the vision of history which Tolstoy expounds in certain reflective passages in War and Peace, its disjunction with the character of Tolstoy’s actual narrative, and the way in which writings by the Counter-Enlightenment Savoyard philosopher and diplomat Joseph de Maistre can be seen as “a direct source and a most arresting prototype” (Reference BerlinBerlin 1951, 19 n1) for Tolstoy’s vision.

The fox knows many things, because it is a beast of multifarious cunning, and the hedgehog knows one big thing, which is to curl into a scarcely penetrable ball of prickles when it is in danger. Tolstoy, said Isaiah Berlin, “was by nature a fox, but believed in being a hedgehog” (Reference Berlin1951, 18): he “perceived reality in its multiplicity, as a collection of separate entities round and into which he saw with a clarity of perception scarcely ever equalled, but he believed only in one vast, unitary whole” (40). The analogy with Tolstoy’s younger contemporary James Murray – their dates are close, 1828–1910 and 1837–1915 respectively – should be evident. On the one hand, the individual entries in the Oxford English Dictionary which Murray did so much to create and edit are dazzling portraits of single words, which speak eloquently for themselves in the quotation paragraphs – and as for his definitions, well might it be said of Murray that “he saw the manifold objects and situations on earth in their full multiplicity; he grasped their individual essences, and what divided them from what they were not, with a clarity to which there is no parallel” (of course, it was not originally said of Murray but of Tolstoy: Reference BerlinBerlin 1951, 38). On the other hand, Murray’s dictionary, like Tolstoy’s novel, was driven by a conflicted belief in a unitary whole: “the reality of inexorable historical determinism” (Reference BerlinBerlin 1951, 33) for one, and the reality of the English language as a single entity for the other. Berlin saw Tolstoy and Maistre as “sharp-eyed foxes” seeking to be hedgehogs, seeking the “single great vision” in vain (Reference Berlin1951, 53). Murray was a wiser fox, prepared to accept a burden of labor which could only be managed by pretending to be a hedgehog, and prepared to work with a single great vision of the English language without denying its incompatibility with the multitude of observable data:

These words could only have been written by a person who knew many things, and whose eye for a draftable entry – and for its converse, a word for which it would be a mistake to draft an entry, because it was so narrowly regional, so closely confined to the slang of an in-group, so obscurely technical, or so poorly assimilated into English – had become very practiced, and hence very sharp. Murray remarked explicitly in the same passage that the taxonomic difficulty at the heart of the dictionary had parallels in nineteenth-century zoological and botanical taxonomy, and when we say that it also had parallels in the internal energies of War and Peace, we are feeling our way toward a question for the intellectual historian about grand theory and the observation of particulars in the world in which Tolstoy and Murray, and their contemporaries in the natural sciences, did their thinking.

So, the specific content of dictionaries may be of interest to intellectual historians. But, at a rather more general level, the structure of dictionaries may also call for their attention.

21.4 Begriffsgeschichte: The Lexicographer as Intellectual Historian

As well as being subjects for intellectual history in both of these ways, dictionaries may be the texts in which intellectual historians express their findings. If we wanted to see Sir Henry Spelman as an intellectual historian avant la lettre, we could take him as a first example; but that might be stretching a point.

A better way in might be to observe that the kind of intellectual history called Begriffsgeschichte (sometimes translated, though not to universal satisfaction, as “conceptual history” or “the history of concepts”) was defined by one of its leading practitioners, Reinhart Koselleck, as “a specialized method of source criticism which attends to the use of terms of social or political relevance.”Footnote ³ Calling a dictionary a work of source criticism feels a little strange until we recapture the likeness between dictionary and commentary which led the humanist Niccolò Perotti to construct a great Latin dictionary (1489) in the form of a commentary on Martial, and led a humanist of the following century, Guillaume Morel, to identify his dictionary of Latin, Greek, and French (1558) as Commentarii rather than as Lexicon or Dictionarium or Thesaurus. But attending to terms is certainly what lexicographers do, and indeed, Koselleck was a co-editor of the nine-volume Geschichtliche Grundbegriffe: historisches Lexikon zur politisch-sozialen Sprache in Deutschland (“Basic concepts in history: A historical lexicon of political-social language in Germany,” 1972–2004; discussed in Reference TribeTribe 2016). We might say that, although it is alphabetically ordered, the very long entries in this work – for instance, Herrschaft “authority, domination, rulership” is over a hundred pages – make it less a dictionary than a collection of essays. The obvious English-language analog would be Raymond Williams’ Keywords (Reference Williams1976 / 1985), which is alphabetically arranged from aesthetic and alienation to western and work.

Nobody accuses Williams of being a lexicographer. But they might, almost. The first draft of Keywords was meant to be an appendix to Williams’ Culture and Society, and that was a book which took its initial energy from Williams’ encounter with a dictionary:

Keywords itself was, he pointed out in the introduction to its second edition, “not a series of footnotes to dictionary histories or definitions of a number of words” (15). Why did he think that anyone would see his book that way? If we take his treatment of jargon (which is around the midpoint of Keywords, the place in an alphabetized text where its modus operandi is likeliest to be found at its peak of development), we see why. After stating that the word “has been in English” from the mid-fourteenth century and is from Old French jargon “warbling of birds, chatter,” Williams quotes six examples of the word in early use. The date of first use, the etymology, and all six examples, are all straight from the OED entry jargon, written by Murray and published in 1901. One of the examples, by the way, is interpreted correctly by Murray and incorrectly by Williams: when a Protestant controversialist wrote about what “the Romanists vnderstand by this Iargon” (Reference BedellBedell 1624, 66), he was not referring to “the terms of an opposing religious or philosophical position” (Reference WilliamsWilliams 1976/1985, 175; emphasis added), but to arguments which, although they were expressed in terms to which he had no objection, struck him on their intellectual merits as “nonsense, gibberish” (OED, s.v. jargon n.¹, sense 3).

What Williams added to his OED material was an overarching argument with present-day reference, which is tightly enough unified to be somewhat misrepresented by excerption. But I shall excerpt anyway: he notes, for instance, that “It is interesting that it is mainly in psychology and sociology, and studies derived from them, but also in relation to an opposing intellectual position such as Marxism, that some of the most regular dismissive uses of jargon are now found,” and, in conclusion, that “To run together the senses of jargon as specialized, unfamiliar, belonging to a hostile position, and unintelligible chatter is then at times indeed a jargon: a confident local habit which merely assumes its own intelligibility and generality” (Reference WilliamsWilliams 1976 / 1985, 175 and 176). That sort of discussion is adumbrated in Murray’s OED entry, which notes that the word is “Often a term of contempt for something the speaker does not understand” and can be “applied contemptuously to a language by someone who does not understand it” (OED, s.v. jargon n.¹, senses 3 and 5), but what Williams offers is indeed, as he protests, more than a footnote to OED. One might say that in Keywords, the foundation of intellectual history which Williams provides is heavily indebted to the dictionary, but that the ideas for which one reads his book are much more independent.

“At present, scholars who call themselves intellectual historians […] can be found working on the history of identity, time and space, empire and race, sex and gender, academic and popular science, the body and its functions, the history of attitudes to food, animals, the environment and the natural world,” and many other topics (Reference WhatmoreWhatmore 2016, 19), and the entries in a large general historical dictionary include more or less explicit histories of all of these. In the case of the OED, bridges are thrown from entry to entry in the OED-derived Historical Thesaurus of English (second edition, 2020); Raymond Williams cross-references from jargon to dialect, but the Historical Thesaurus arranges the materials for a whole history of disparaging attitudes to other people’s language, from jargon (which has the earliest first attestation of any word in its semantic set, 1340–1370), gibberish (circa 1557), and fustian (no later than 1593), all the way to mumbo-jumbery (1923), mumbo (1931), and double-talk (1938). A further project, “Mapping Metaphor with the Historical Thesaurus” (for which see Reference Alexander, Anderson, Bramwell and HoughAlexander 2016), penetrates even further into the intellectual historian’s question of “what people in the past meant by the things they said.”

More specialized historical dictionaries focus more closely on questions of intellectual history. Eminent among them, at least in a western European context, is the Deutsches Rechtswörterbuch (“German Law Dictionary,” 1912–present), which documents the legal vocabulary of German, broadly conceived. Its chronological range extends from late antiquity to the nineteenth century, and it presents abundant early material from other Germanic languages including Old English. A bibliography appeared in 1912, and the first volume began publication in 1914. The thirteenth was published in 2018, bringing the total number of published entries to 97,000, many of them substantial: sprechen ‘speak,’ for example, has thirty-two main senses (other examples of elaborate entries are noted in Deutsches Rechtswörterbuch, vol. 13, iii–iv). Just as OED material reappears in Raymond Williams’ Keywords, so material from the Deutsches Rechtswörterbuch has been treated both in dictionary form and also in continuous prose, an attractive example of the latter being the use of the unpublished raw materials for a future entry Zigeuner (denotatively ‘Roma,’ but connotatively the word has often been used similarly to English Gypsy) in an article on representations of Zigeuner in texts bearing on the law (Reference Kronauer and AwosusiKronauer 1998).

In the Deutsches Rechtswörterbuch, dictionary materials shed light on questions of collective intellectual history – “what,” they help us to ask, “did German-speakers mean when they said Zigeuner, and what did the category of ‘Zigeuner’ mean to them?” – and in another major German dictionary, it is the history of a single intellect which is illuminated. This is the Goethe-Wörterbuch, work on which was launched in Berlin at the end of 1946 (what indomitable courage it must have taken to begin such a project in that city at that moment!), and which had by the end of 2019 treated the vocabulary of Johann Wolfgang von Goethe, whose works run to 143 volumes in the standard edition, from A to radikal, in six completed volumes and the first fascicle of a seventh. Among the first specimen entries to be drafted at the beginning of the project were Licht ‘light’ (an important word for anyone, but particularly for Goethe, who published on optics); sammeln ‘to collect’; and schön ‘beautiful’ (Goethe-Wörterbuch, vol. 1, 5): so, there was a clear intention, from the beginning, to make the dictionary a portrait of Goethe’s thought. A Shakespeare dictionary on such a scale would be the best intellectual biography of Shakespeare which could be imagined.

Molly Bassett remarks (Reference Bassett2015, 47) that “[t]he interdisciplinary nature of Mesoamerican studies makes distinguishing lexicographers from art historians from scholars of religions rather artificial,” and in so far as the studies in question are historically oriented, one might say that the lexicographers, like the art historians and the scholars of religion, spend part of their time doing intellectual history. This is true far beyond the borders of Mesoamerican studies.

21.5 Noli Altum Sapere: Dictionaries and the History of Intellectual Institutions

So, first, intellectual historians may study dictionaries for the sake of their content; second, they may study them for the sake of their form; and third, intellectual histories may take dictionary form or may approximate it. Those are three avenues of inquiry into the relationship of dictionaries and intellectual history. Now let us follow a fourth, by pursuing the topic of the relationship between the history of dictionaries and the history of intellectual institutions and communities. In one formulation, we are now moving from intellectual history as “the history of thought,” which is expressed in “works that focus on the history of ideas, language, texts, ideology, meaning and cultural representation,” to intellectual history as “the social history of intellectuals,” which includes “intellectual biography, histories of institutions (universities, salons, reading groups, professional organizations), publishing, authorship, and reading” (Reference WickbergWickberg 2001, 384). We can do that without moving too far into the domain of book history, which is the subject of another chapter in the present volume.Footnote ⁴

21.6 The Global Intellectual History of Dictionaries

“The departmentalization – whether by subjects, periods, nationalities, or languages – of the study of the history of thought corresponds, for the most part, to no real cleavages among the phenomena studied,” as A. O. Lovejoy remarked in the introduction to the first issue of the Journal of the History of Ideas (Reference Lovejoy1940, 4). “Ideas,” he concluded in a much-quoted sentence, “are the most migratory things in the world” (4). Lovejoy’s immediate point was about the interdisciplinarity of that part of intellectual history called the history of ideas. But his words have been used to make a point about the international, indeed the global, quality of the field, for instance by David Armitage in a discussion (Reference Armitage2013, 17) of “the international turn in intellectual history.”

Nobody who has thought seriously about the practice, or the history, of lexicography could fail to see what an international business it is and has always been – hence the recent publication of a Cambridge World History of Lexicography (Reference ConsidineConsidine 2019d). An aim of this chapter has been to suggest that lexicography is a very suitable topic for an international, even a global, intellectual history. Here, the flexibility of the concept “lexicography” is helpful: we already know that words have been gathered in many different genres at different times and in different places, and that the kind of wordlist Henry Spelman made was very unlike the wordlists his contemporaries in other parts of the world were making. Perhaps the concept “dictionary” is less flexible, and there have been cultures with flourishing lexicographical traditions in which there was no one word equivalent to modern English dictionary (see, for instance, Reference VedalVedal 2019, 110–111, and Reference Vogel and GondaVogel 1979, 303n1). But a global intellectual history of lexicography need not impose a concept proper to one culture on the description of other cultures.

What it might do, conversely, is ask how lexicographical ideas traveled from one culture to another. As far as I know, there is no reason to doubt that the first Chinese lexicography was entirely independent from the early lexicographical traditions of Mesopotamia and the eastern Mediterranean. On the other hand, Chinese lexicography certainly received some influence from Sanskrit traditions of the analysis of language, transmitted by Buddhist teachers, during the first millennium AD (Reference VedalVedal 2019, 127–128), and it was, in its turn, the source of the lexicographical traditions of Japanese, Korean, and Vietnamese (Reference Söderblom SaarelaSöderblom Saarela 2019). Again, on the one hand, the first Arabic lexicographers were surely not unaware of the earlier eastern Mediterranean lexicographical traditions of Greek and Coptic. And on the other, the inspiration of the first Hebrew dictionary was Arabic (Reference MamanMaman 2019, 186–187), and the first Latin dictionary with the strongly derivational macrostructure which is to be seen in Arabic dictionaries of an earlier date was made at a time when the high culture of Latin Christendom was responding to cultural influences from the Arabic-speaking world (Reference Considine, Benati and HändlConsidine 2019c, 115–117). A fully developed global history of lexicography will take account of the ways in which lexicographical ideas have migrated.

“Quite what a global intellectual history would comprise, or even what its subject-matter will be,” writes Armitage, “is still far from clear” (Reference Armitage2013, 32). But that global intellectual history is where the intellectual history of lexicography, at its most ambitious and most exciting, ultimately belongs.

22.1 Introduction

How are dictionaries shaped by social history, and how far do dictionaries themselves shape social history? Established texts reflect particular perspectives – the earliest wordlists in England reflect for example the world views of male clerics learning Latin, and later of aristocratic women teaching their children vernacular terms relating to animals and agriculture. Wordlists and dictionaries (broadly defined) are also influenced by tradition but may be adapted for new audiences. As the instruction of boys extended beyond the church to town schools, topical glossaries added more words pertaining to women. After meat eating distinguished the secular upper classes from everyone else in the medieval world (Reference EliasElias 2004, 11–12), carving terms for fish, fowl, and meat spread from manuscript manuals for medieval leisured readers into eighteenth-century publications aimed at female servants. No eighteenth-century servant would use the phrases “Culpon a Trout” or “Transon that Eel” (M. Reference JohnsonJohnson 1755, xii), but an ambitious servant might recognize them. Archaic texts and text types might in turn inspire innovation: words like culpon and transon would be defined in what was to become the Oxford English Dictionary (OED), as history became a defining index of national identity just as the British empire spread English around the world. As well as containing information, dictionaries perpetuate ideologies.

This chapter maps the most significant historical intersections of English dictionaries and Anglophone societies. It spans the shift from English as a colonized to a colonizing language, from the medieval period to around 1900. Its building blocks include intersecting conceptions of gender roles, the family, social status, work and industrialization, as well as urbanization and racialization, which are some of “the fundamental elements of social life,” according to Britain’s History Workshop Journal’s Editorial Collective (1976, 1). Some other concepts remain implicit in this chapter. Education interconnects every section. It was in religious contexts that Latin was codified, and methods were perfected for organizing words within books as well as books within libraries (Reference ClanchyClanchy 2013, 183; Reference FlandersFlanders 2020, 27–130). The idea of the nation was later shaped by the OED with history and by the state with nineteenth-century mass primary education. Overall, what the British social history tradition describes as “tension between human agency and […] determinism” is brought constantly into the foreground (Reference Morris, Crowcroft and CannonMorris 2015). My focus on English lets me demonstrate my expertise and contrast revisions of the “same” text within the limits of a handbook chapter. My approach is anecdotal: I relate social changes to identifiable revisions and initiatives by individual lexicographers.

I consider lexicography and define dictionary broadly. Collections of ordered words and terms predate the existence of the Latin word dictionarius. Wordlists might be structured in rhyme or at random and appear in other genres – versified vernacular vocabularies for motherly instruction of future land magnates or terms for carving or collective nouns that persist through generations of courtesy books, a genre that initially advised aspiring readers on how to behave at noble courts. Dictionaries are not always distinguishable from other genres. What is the status of industrial words when general dictionaries are simultaneously distinguishing themselves from encyclopedias but industrialization is becoming integrated in every Briton’s life? I also touch on how dictionaries are produced and distributed. In library sale catalogs and in mail order catalogs, dictionary size and language suggest a hierarchy of price and status, revealing not only their social stratification but also the industrial networks that enabled their dissemination. I am particularly attentive to how authors revised popular texts in reaction to cultural changes, and how their representations of marginalized humans might reflect those changes and their often glacial progress. Representations of categorized humans can appear as headwords, in quotations or labels, or as an assumed audience. Interpretation is sometimes tricky: in a medieval noble’s household manual, which instructs on managerial as well as practical tasks, does the collective noun the multiplieng of husbondis (Boke of St Albans 1486, f7r; see “Berners, Juliana”) suggest “the procreative enthusiasm of the male” or “the loose morals of married women” (Reference HandsHands 1975, 85 and 161)?

Domestic reference works – the kind that define terms in the context of describing cooking, etiquette, medicine, or other household tasks (Reference RussellRussell 2018, 118) – arbitrarily anchor each section. Their content and audience shift along with intersecting conceptions of family and of social status, as we see from their cataloging in Lindsay Russell’s generative survey of Women and Dictionary Making (Reference Russell2018). As literacy expanded, so did audiences, and spelling dictionaries became common. A dictionary of Christian names within an 1883 Toronto cookbook underscores the conservatism of settlers as they replicated themselves with personal names from classical and Germanic traditions (Reference ClarkeClarke 1883). Persistent carving terms and increased culinary terms suggest the downward diffusion of gentility along with literacy, as well as the outward expansion of empire.

The conclusion summarizes the chapter from a more form-oriented perspective. Dictionaries both reflect and shape these shifting and integrated social categories and the humans who embody them, in combinations of content, production, distribution, and impact. Dictionaries primarily function to facilitate communication by identifying and bridging lexical differences between different social groups. Some dictionaries thus also represent and sometimes recognize or renovate shifting social categories.

22.2 Gender

In pre-modern England, dictionaries reflect male domination of religious and educational institutions as well as the inferior status of the vernacular in many domains. But at least one early wordlist had a conventionally feminine muse. “Grammatica” is credited with some of the content of Osbern Pinnock’s twelfth-century Panormia or Liber derivationum (c. 1148–1179). In its prologue, she appears in a dream and announces her intent to pour lists of words “into the thirsty openings of his memory” (Reference 687Osbern and BusdraghiOsbern 1996, I: 8, translated by Reference Considine, Benati and HändlConsidine 2019c, 112). Osbern was appropriating the authority of the “mother of Latinity” to authorize his innovative attention to derivational morphology (114). A Benedictine monk, Osbern was male, like all the other Britons definitively listed by the Oxford Dictionary of National Biography (ODNB) as authors of medieval wordlists. These manuscripts codified Latin and were targeted to males, if not always explicitly. John of Garland’s Dictionarius (c. 1218, revised c. 1230), which provides seemingly the first recorded use of the word dictionarius, was written by this university teacher pene puer pueris “for boys though almost a boy himself” (Reference LawlerLawler 2006). An author’s intentions did not necessarily restrict his audience: Geoffrey the Grammarian’s Promptuarium parvulorum (1440) means “a storeroom for young scholars.” But although it was “nominally intended for novices,” this pioneering friar’s English–Latin dictionary was used more widely (Reference BurnleyBurnley 2004), especially with the “expansion” of “education through private philanthropy” from religious to town schools from the late fifteenth century (Reference Lawson and SilverLawson and Silver 1973, 44; Reference Orme, Trapp and HellingaOrme 1999, 454). The small girls from wealthy burgess families who attended schools in “a few towns” likely learned prayers in Latin and English, but it was boys who might advance and actually learn the language (Reference Orme, Trapp and HellingaOrme 1999, 451). While the extent of minimal literacy in Latin should not be underestimated, the language allied its learners with Christianity and Christendom (Reference ConsidineConsidine 2019b, 267–268). And especially after the Norman Conquest, the limitations of English might be illustrated by the fact that John of Garland wrote in Latin and taught in France.

One medieval wordlist aligned women with vernaculars. Walter of Bibbesworth’s Tretiz (c. 1240–1250), a classified verse vocabulary in Anglo-Norman, was directly addressed to an aristocratic mother, Denise de Munchensi (d. 1304) (Reference HuntHunt 2005). It contained lists of animal cries, as well as collective nouns for animals, some glossed in English (Reference SayersSayers 2009, 526): for example, “Chat(e) mimoune ‘mewith’, cerpent cifle ‘cisses’” (see MED sissen ‘to hiss,’ Walter de Reference Bibbesworth and RothwellBibbesworth 2009, 11, line 254). The Tretiz also instilled the practical literacy required by prospective “lords of seignorial estates,” as nouns and verbs like blez muez ‘stack grain’ epitomize “the technical language of husbandry and estate management” (Reference 727Keiser, Trapp and HellingaKeiser 1999, 472–473): “En grange vos blez muez ‘mouwe’ / Dehors la grange vos blez tassez” (Walter de Reference Bibbesworth and RothwellBibbesworth 2009, 16, lines 350–351). With the dedication to “madame Dyonise de Mountechensi” (Walter de Reference Bibbesworth and RothwellBibbesworth 2009, 1) remaining in six of the sixteen surviving manuscripts (Reference HuntHunt 2005), the Tretiz’s popularity through the beginning of the fifteenth century perpetuated its representation of maternal instruction as well as demonstrating the utility of rhyme as a memory aid. Containing some glosses in English (e.g. mouwe and mewith), this text also suggests aristocrats’ growing reliance on English as well as women’s general restriction to the vernacular (Reference Franzen and FranzenFranzen 2012, xxxi–xxxii). While some aristocratic women were well educated at nunneries, the books that remain from women’s foundations reflect more reading in English or French vernaculars, except for Latin liturgical books (Reference Baswell and WallaceBaswell 1999, 144–146). By 1400 it seems to have been English that “most daughters of the nobility and gentry” learned to read, whether in nunneries or in their own households (Reference Orme, Trapp and HellingaOrme 1999, 451). Clanchy asserts “it is probable that [mothers] had always been involved” in teaching their children to read, although dictionaries do not appear among his “better evidence” for motherly vernacular literacy instruction in medieval times (Reference Clanchy, Leyser and Smith2011, 129).

Did wordlists of Latin reflect gender roles in England? Texts classifying vocabulary by topic tell us something about social categories, although their compilers were indebted to tradition and unconcerned with women (Reference HüllenHüllen 1999, 173 and 192). Yet small innovations seem suggestive. As the instruction of boys extended beyond the church to town schools, twelfth-century treatises like De utensilibus ad domum regendam by Adam of Balsham (1100/1102?–1157–1169?) expanded to incorporate “objects of everyday life” including “implements peculiar to women” (Reference Franzen and FranzenFranzen 2012, xxiv–v and xli; Reference HüllenHüllen 1999, 68–71; Reference LendinaraLendinara 1999, 377). John Withals’ popular printed English–Latin Dictionarie (1553) exemplifies how topical dictionaries represented the contemporary “world of women” (Reference HüllenHüllen 1999, 191–201). Withals’ words for family relations show women’s distinctively passive role in marriage: “He that marieth ‘sponsus, si’, / She that is married ‘sponsa, sae’” (1553, 70^r; Reference HüllenHüllen 1999, 200). Words in the masculine gender dominate lists of public activities and occupations, although the author must have known in his capacity as a Latin teacher that “biological and grammatical meanings often do not coincide” (Reference HüllenHüllen 1999, 197); historical records show that women worked in more spheres than Withals’ dictionary represents (Reference HüllenHüllen 1999, 198). The few feminine-gendered nouns represent women workers in agriculture, cloth, food, and laundry – and as heads of households (Reference HüllenHüllen 1999, 198–199). The dictionary’s over-representation of clothing for men also illustrates clothing’s role in signifying status: women’s was defined by their marital state in contrast to men’s in the traditional “feudal hierarchy of those who fought, prayed, and worked” (Reference HüllenHüllen 1999, 199). This popular text retained its structure as well as its influence throughout almost eighty years and over multiple editions (Reference HüllenHüllen 1999, 169–170). Hüllen argues that this Shorte Dictionarie’s representation of the world “from above to below” and “from outside to inside” must have been appropriate for teaching Yonge Begynners because it corresponded to how humans perceived reality (Reference Hüllen1999, 179–181).

A wordlist giving hints at women’s social roles appears in the so-called Boke of St. Albans (1486), a frequently reprinted guide to hawking, hunting, coat-armor and heraldry (Reference HandsHands 1975, xiii). Parts of the Boke have been controversially attributed to an individual who might have been female and who (if so) would therefore qualify as England’s first-known female “lexicographer.” The “dame” seemingly responsible for a few of its pages (d3^v–4^r, f4^v–8^r) has sometimes been identified as “Juliana Berners” – spelled in various ways in different editions (Reference BoffeyBoffey 2004; Reference HandsHands 1975, xiv–xv and lv–lx). Regardless of the sex or identity of this J. B., the compiler of the book seemingly believed that his authority was a woman, with whatever social position the word Dam accorded in the fifteenth century (Reference HandsHands 1975, lix; OED3, dame, n. II 5). Some material now unlikely attributable to J. B. features a dialogue between a mother and son in the art of hunting, showing women’s roles as mediators of male knowledge and the importance of “classification and nomenclature” (Reference BarrattBarratt 2010, 250). Indeed, the manual popularized not only the “practice” but also “the language proper” to the subjects (Reference HandsHands 1975, xiii).

Conventional gender stereotypes appear in a peripheral but persistent semi-lexicographical genre, a list of collective nouns found in thirteenth-century texts like Bibbesworth’s (Reference Scott-MacnabScott-Macnab 2003, 83). In the Boke of St. Albans, this “Compaynys of Beestys and Fowlys” duly follows the sections on hawking and hunting and has been associated most definitively with “J.B.” (Reference HandsHands 1975, xiv and 152; Reference Scott-MacnabScott-Macnab 2003, 4 and 43). Of potential relevance to social history are phrases like “a promyse of tapsteris” and “a scolding of kemsteris” [‘combers’] (1486, f7^r). Were the kemsters and tapsters biologically as well as etymologically feminine? It is unclear whether combing and scolding were connected via bad temper (Reference Scott-MacnabScott-Macnab 2003, 285) or possibly gossip (e.g. MED, scōld(e, n. 1a); scolding was not necessarily a feminine mode of speech (OED3, scold, v. and scolding, n.). And if tapsters were female, then “the significance of ‘promise’ as the collective becomes open to several interpretations” (Reference HandsHands 1975, 162–163). However, a tapster eventually referred to both alewives and males (Reference Scott-MacnabScott-Macnab 2003, 302). And that Tapster is classified as a “surname” (MED, tappester(e, n. 1c) for names ranging from John le Tappistere (1317) to Iohanna Tapstere (1390) also illustrates the shift from bynames including occupational terms to family names through the Middle English period (Reference Clark and BlakeClark 1992, 566–583). How medieval names were ordered (within concordances or catalogs or government records) is another, long story: “Medieval intellectuals were ambivalent towards alphabetical order because it militated against their sense of hierarchy” (Reference ClanchyClanchy 2013, 182). Since names were often listed for such administrative purposes as taxation, women’s names were generally omitted altogether from medieval Poll Tax Rolls, “which do often name servants and other subsidiary adult members of households” (Reference Clark and BlakeClark 1992, 545).

22.3 Family

Concepts of family (structure and roles) intersect with social status – in medieval (and all other) times. In the fifteenth century, the word family could denote the servants of a household (OED3, family, A I 1ab), a group of people living together including servants or boarders (2a), as well as people related to each other, whether living together or not (2b). The compilation and printing of the Boke of St. Albans (1486) brings the noble household into focus – along with the shifting concepts of nobility and of family. The Boke’s supposed contributor or compiler, “Dam Julyans Barnes,” was a plausible “lexicographer” because in this and other texts aristocratic women represented appropriate transmitters of noble values. The Boke has been characterized as “a work of noble education” because of its focus on the terminologies of hawking, hunting, coat-armor and heraldry (Reference Orme, Trapp and HellingaOrme 1999, 458); it resembles texts targeted earlier by Caxton to support the “vernacular education” in noble households (Reference Orme, Trapp and HellingaOrme 1999, 456). Its terminologies for breekying or dressyng of dyuerse beestis and fowlis ‘carving’ (1486, f7^v) connect hunting and hawking with the “rituals of dining” that dominate “the service discussed in conduct texts” more generally (Reference 742PinyanPinyan 2017, 205): “A dere brokenne,” “a cony vnlaciedde” (1486, f7^v). For young male readers of contemporary conduct poems, ancestry was only one of the “main paths to gentility” (Reference 742PinyanPinyan 2017, 78 and 86 and 203–205). “[T]he service of a page boy or henchman to his lord” was another criterion of gentility in contemporary conduct literature (Reference 742PinyanPinyan 2017, 204); the Boke reminds readers about service in noble households of all levels and kinds, instructing readers how to care for hawks by feeding or medicating them (1486, a3^v) while randomly mentioning servants, ushers, butlers, and stewards in its list of collective nouns (“a prouision of stewardis of hous” f7^r), the steward heading the household staff hierarchy (Reference Scott-MacnabScott-Macnab 2003, 300). The association and commodification of household service with rituals of dining can also be seen in the verse Boke of Nurture apparently compiled by John Russell, formerly in service with Humfrey, duke of Gloucester (1391–1447) as “usher in chamber and marshal in hall” (Reference GrayGray 2004). This manuscript would be abbreviated by Caxton’s successor Wynkyn de Worde and printed twice as The boke of keruynge (1508, 1513) “and repeatedly thereafter” (Reference 727Keiser, Trapp and HellingaKeiser 1999, 484–485). A medieval nobleman’s ability to carve a whole animal at the table indicated the size and the wealth of his household (Reference EliasElias 2004, 11–14).

To some scholars, the fact that terminology accompanied instruction suggests how gentility was commodified and disseminated beyond the noble household and indeed beyond the traditional category of landholders. The Boke drew on manuscript traditions that include fifteenth-century miscellanies whose contents and ownership suggest a readership of “the English landholding classes and those responsible to them” (Reference 727Keiser, Trapp and HellingaKeiser 1999, 471–472). But curiously, its nameless printer otherwise published Latin scholarly texts: indeed, England’s third printer was later identified as a schoolmaster (Reference HandsHands 1975, xv–xvi). Scott-Macnab identifies other owners of manuscripts associated with “the J. B. material” in the Boke as “stretch[ing] far beyond the traditional hawking-and-hunting landed classes,” including a “city mayor, a schoolmaster and a clergyman” (Reference Scott-Macnab2003, 88). Scott-Macnab claims that the Boke “differs from most other practical books, and all the other English treatises on hawking and hunting, in that it provides information about the terminology of these sports, but offers little or no instruction concerning their actual practices” (Reference Scott-MacnabScott-Macnab 2003, 87). While Kristin Pinyan agrees that the manual focuses on language as well as skills, she also observes that English hunting manuals are generally more attentive than Continental texts to terminology, and one context for their dissemination is the expansion of English as a language of gentility (Reference 742Pinyan2017, 327 and 329–339).

Within household manuals, including and organizing wordlists links shifting concepts of the household with literacy and social mobility. As authors of household manuals, women might index both family status and domestic authority. Like the name of “Dame Juliana Berners” in the late fifteenth century, the name of Hannah Wolley was misleadingly linked with several seventeenth-century domestic texts containing a few wordlists – carving terms again, as well as terms of address to be used in letter-writing (Reference RussellRussell 2018, 73–74 and 118–120). The author Wolley existed (1622?–1674 or after). Her life as an elite household attendant and later as a grammar-school teacher’s wife prepared her for widowhood and authorship “on cookery, medicine, and household affairs” (Reference ConsidineConsidine 2004). While Wolley’s acknowledged works do not contain wordlists, their organizational devices link literacy with labor rather than leisure: the alphabetized table of contents for The Ladies Directory contains multiple entries for “Cake” and “Consumption” (Reference WallWall 2010, 394; Wolley [1661] 1662, A5^v–A6^r). Such publications made elevated life more legible for readers of lower rank as literacy increased in society (Reference EllisonEllison 1999), although their titles addressed specific audiences. Of her authorized publications, The Queen-Like Closet (1670) “reused and added to material from her two earlier books,” The Ladies Directory (1661) and The Cooks Guide (1664; see Reference ConsidineConsidine 2004).

Some relations between household structures and wordlists might be inferred from texts that were attributed to Wolley without authorization. Their socially stratified titles distinguish The Gentlewomans Companion (1673) and The Compleat Servant-Maid (1677), as does the assumption that only the Gentlewoman will use a (short) list of “proper titles” for addressing letters (1673, 226). The Gentlewomans Companion distinguishes stratification within the family: “If kindred write one to another, the greater may express the relation in the beginning of the Letter; but she that is of the meaner quality, must be content to specifie it in the Subscription” (1673, 227). Both texts also attest to the shifting impact on the household of the growing wealth of the rising commercial and professional classes. Wives hired servants and thus might lose or devalue “the traditional female skills of running the household” (e.g. Reference EllisonEllison 1999, quoting Hobby 1989, 171). The Gentlewomans Companion instructs female employers how to manage “unfaithful” servants (1673, 109) and how to supervise laboring ones. Indeed, the ubiquitous carving terms (1673, 113) directly follow a section explaining that “understanding how to dress Meat as well as eat it” is “in no wise dishonourable […] that your Servants may be guided by you, and not you by them” (1673, 112). The carving terms in both texts (1673, 113; 1677, 30) remind us that leisure and labor intersect at the dining table. While literacy is difficult to define and measure, the existence of self-help manuals made terms and skills visible to ambitious servants and to the leisured wives who supervised them. Wolley had represented the medical learning from her mother and sisters as greatly amplified by doctors and by books during seven years in service before marriage (1674, 10–11). The Compleat Servant-Maid specifically teaches its readers “how they may fit, and qualifie themselves for […] Employments” such as “Waiting-Woman” and “House-keeper” (1677, t.p.), and if “a Complete Servant-maid” then also a “Mistress” and “Wife” (A5^r –A5^v). While Wolley described herself as “belong[ing] to a Noble Lady” before marriage (1674, 10), this publication in her name promises its purchasers the prospect of autonomy.

Over time, servants became excluded from household ties and from concepts of the family. But linguistic evidence of what Tadmor calls the “household-family” persists in dictionaries and other sources through the eighteenth century, when “the nuclear family” was not yet “the abiding organisational and cultural epitome of domestic and familial relationships in seventeenth- and eighteenth-century England” (Reference TadmorTadmor 2001, 103–166). In what its modern editors have called “the first substantial reference book to be published in England with women as its principal target audience” (Reference Considine and BrownConsidine and Brown 2010, vii), The Ladies Dictionary (1694) defines the “Family Duties” in marriage as inevitably including “Children and Servants” (N. H. 1694, 347). Later, in the wake of Linnaeus, botanical dictionaries used by women presented other images of natural families: Charlotte Murray describes “the Generic name of a plant” as “the family sirname” [sic] in the “Introductory Remarks” to her British Garden (Reference Murray1799, I: vii). Perhaps because of women’s long association with domestic medicine, botany became not only a gender-neutral but a feminized field (Reference RussellRussell 2018, 110–111).

Roles and responsibilities within the nuclear family are suggested by the authorship and the content of wordlists. Women being ideally confined to the household and necessarily dependent upon such male family members as brothers is exemplified by the attribution of The Family Magazine to “Arabella Atkyns” (1741). The collection is presented as a systematization of her deceased physician brother’s “Common-place Book” (vi) compiled alongside “the Subjects of Family Management” (x) but presented pseudonymously on the grounds of her “Relation to several honourable Persons” (xiv; quoted by Reference RussellRussell 2018, 74–75). As well as the inevitable “Terms of Art used in Carving” (1741, iv; 19–20), we find “difficult Terms that occur not usually in common Reading […] in a short kind of Dictionary at the End” – from Abdomen to Uterus (ix; 316–318).

Women and their instructional roles are (made) invisible in the anonymous and privately printed Practical Education (1780), apparently co-authored by Richard Edgeworth and his wife Honora before her death that year. In it, glossaries “give […] clear and accurate ideas of every word” (1780, iv–v) – of eighty-five of the words used in the dialogues in the text (Reference Iversen, Haugland, McCafferty and RustenIversen 2014, 340). Women’s roles as domestic educators become more explicit in the textual tradition that followed. Honora’s stepdaughter Maria expanded the glossary from the 1780 text in her own Early Lessons (1801) (Reference Iversen, Haugland, McCafferty and RustenIversen 2014, 241); she became an influential author in her own right after educating her half-siblings and collaborating with her father as an author. And in turn Anna Brownell Murphy (later Jameson) acknowledged the influence of “Miss Edgeworth” and her “little glossary” ([c.1813], iii); the title of her The First or Mother’s Dictionary for Children made explicit mothers’ longstanding obligations to talk their children into literacy, and “some of the entries […] read more like conversations than dictionary definitions” (Reference Iversen, Haugland, McCafferty and RustenIversen 2014, 343). Murphy published her Mother’s Dictionary while employed by the marquess of Winchester (Reference JohnstonJohnston 2004), demonstrating the governess’s liminal social role.

22.4 Social Status

By classifying vocabulary and disseminating it to particular audiences, dictionaries reflected and perhaps stretched social hierarchies, in domestic and more public domains. “Ladies, Gentlewomen, or any other vnskilful persons” featured in the audience for Robert Cawdrey’s A Table Alphabeticall (1604). Sylvia Brown argues that Cawdrey, a Puritan clergyman, believed that literate Protestant women would maximize their evangelizing potential as mothers if they knew “hard vsuall English wordes, borrowed from the Hebrew, Greeke, Latine, or French, &c” (1604; 2001, 143). Five specific “Ladies” were dedicatees: the daughters of Lucy Sidney Harington, who was Cawdrey’s “especiall friend in the Lord” as well as the sister of his former “scholler” and present “benefactor,” Sir James Harington (1604, A2^v). In the wake of humanism, many female dedicatees were learned as well as generous. But “stereotypes” can be “mobiliz[ed]” for reasons in addition to representing reality (Reference RussellRussell 2018, 37). Associations with “Ladies” might have flattered “unskilful” men of lower status – though the expansion of grammar schools was expanding the demographic of Latin scholars to “sons of the landed elite and of the ‘middling sort’ of men in the professions and merchant elites” (Reference GreenGreen 2009, 76). The same scholar might have access to a range of publications, as the more differentiated English Latin dictionaries illustrate. At Shakespeare’s Stratford grammar school, Thomas Cooper’s “massive folio” Thesaurus linguae Romanae et Britannicae (1565) was used for reference (Reference GreenGreen 2009, 1–2). But Shakespeare likely also used Withals’ Shorte Dictionarie (1553): this English–Latin vocabulary became “the regular grammar school dictionary for memorization” (Reference BaldwinBaldwin 1944, 531, 709–720). Renaissance pedagogy involved double translation, in and out of “Latins” and “vulgars”: Touchstone’s mock-definition of society “which in the boorish is company” (As You Like It, V, i, 52) might have come from Cooper’s definition of consortium “Felowship: cōpanie: societie” (1565, Dd3^r; all quoted by Reference BaldwinBaldwin 1944, I: 716–717). Another English–Latin dictionary occasionally distinguished “vulgare speech” – Richard Huloet’s Abecedarium Anglico-Latinum (1552) (Reference Stein, McConchie, Timofeeva, Tissari and SäilyStein 2006, 28–30).

Monolingual dictionaries increasingly distinguished not only hard ‘learned’ but low words – and their users. Henry Cockeram aimed at a broader audience than Cawdrey and his successor John Reference BullokarBullokar (1616) – that is, one including not only “Ladies and Gentle women,” but “young Schollers, Clarkes, Merchants” and “Strangers” (1623, t.p.). The second part of Cockeram’s tripartite English Dictionarie (1623) essentially translated “vulgar” headwords with definitions characterized as “more refined and elegant” (A4^v). Later lexicographers in the hard-word tradition refined ways of guiding usage: in 1658, after old words were asterisked, Phillips introduced a dagger for marking excessively Latinate words (Reference OsseltonOsselton 2006, 101). Thomas Blount innovatively cited authorities in his Glossographia (1656) (Reference Considine and ConsidineConsidine 2012b, xxvi) – and perhaps not incidentally was an early codifier of the social sense of class: “an order or distribution of people according to their several Degrees” (quoted by Reference CorfieldCorfield 1987, 47). Social judgments proliferate along with usage labels.

Usage labels in Samuel Reference JohnsonJohnson’s 1755 dictionary include “low,” “barbarous,” “vulgar,” and “cant.” Fun is “a low cant word” (quoted by Reference HudsonHudson 1998, 78), also found in B.E.’s Dictionary of the Canting Crew (1699) before infiltrating respectable society and standard English: for Carey McIntosh, Frances Burney’s eulogy of Johnson himself as endowed with “more fun … and love of nonsense about him than almost anybody I ever saw” suggests the status of that word by 1779 (2020, 141). Of course, imagined audiences and headwords and usage labels in dictionaries do not necessarily reflect “the social life of the English language” (Reference Considine and ConsidineConsidine 2012b, xx). And even for modern corpus linguists, ascertaining “social status variation” of lexemes other than terms of address is difficult because of the relatively low frequency of non-grammatical words and the higher social status of most historical writers (Reference NevalainenNevalainen 2006, 139). But codifiers’ motivations do underscore social issues, such as the importance of primogeniture and of Protestantism in English society – Cockeram was a younger son of Devonshire gentry and possibly a schoolteacher, while Blount was a Catholic recusant with legal training (Reference Considine and ConsidineConsidine 2012b, xxiv–xxvii).

Socioeconomic distinctions might be drawn from dictionary ownership. In the seventeenth century, larger hard-word dictionaries like Blount’s and Phillips’ “competed for the top end of the market,” while Bullokar and Cockeram shared the bottom end (Reference ConsidineConsidine 2012a, 49). In the earlier eighteenth century, non-classically educated speakers were stereotyped as using editions of Nathan Bailey’s popular dictionaries (1721–1727, 1730) as shortcuts to the learning they lacked – not only the agricultural laboring poet Stephen Duck in 1730 but also the fictional Mrs. Western in Fielding’s Tom Jones (1749). Melissa Patterson observes that “at six shillings, the book was more than a week’s wages for a thresher” (Reference Patterson2015, 46n108, 41–56; see also Reference HancherHancher 2009).

Dictionaries appear in private libraries, although sale catalogs are notoriously unreliable guides to book ownership and usage: for instance, in 1731 Daniel Defoe’s library was combined not only with another collector’s but also with additional stock from the bookseller (Reference GehlingGehling 2015, 5). Defoe’s ownership of Bailey’s Universal Etymological English Dictionary (1726) thus cannot be proven, but the classification of Bailey under “English miscellanies. Octavo” shows how classifying books by size and by language creates a hierarchy of publications – though all ostensibly belonging to one person (Reference Payne and HeidenreichPayne 1970, [v], 56, 59). In inventories of nine of the eleven larger private libraries described by Dunstan in late eighteenth-century Scotland, two collectors owned a copy of Samuel Johnson’s dictionary; so too did a brewer (Reference DunstanDunstan 2012, 277 and 282–283). But which edition? At over four pounds, the 1755 folio was particularly expensive relative to the many other English dictionaries published between 1755 and 1775 – over eighty new titles and new editions (Reference 738Mugglestone, Johnston and MugglestoneMugglestone 2012, 141 and 143).

As lexicographers codified the “general” vocabulary as well as hard words, dictionaries proliferated to match broad demographics (perhaps as much aspirational as actual). Johnson’s would be refashioned through the nineteenth century for a variety of audiences (Reference DilleDille 2005). While a condensed octavo edition cost ten shillings in 1756 (Reference 738Mugglestone, Johnston and MugglestoneMugglestone 2012, 148), the domestic reference work called Madam Johnson’s Present (formerly The Young Woman’s Companion; or the Servant-Maid’s Assistant 1753) cost 1s. 6d. but “contains double the Quantity that is usually sold for that Sum” (M. Reference JohnsonJohnson 1754, 5), including “a new English Spelling Dictionary” containing ordinary words – from abate to zone – and proper names (M. Reference JohnsonJohnson 1755, 30–60).

Terms of address were changing through this period: McIntosh links the history of mister with both upward and downward social mobility “during the same period that enclosure and the waning of feudal tenure produced ever-growing numbers of landless male workers” (Reference McIntosh2020, 48–53). Madam was used by those in service to address their mistress (OED3, madam, n. 1a), so although (with or without a word like Honoured) Madam had also become the all-purpose word used for addressing letters to women of any rank (e.g. M. Reference JohnsonJohnson 1755, 81–85), perhaps Madam Johnson’s appearance in the 1755 title emphasized her domestic authority over the social status of purchasers, one of whom was not a servant-maid but the Widow Fox’s manservant (Reference 716Fergus, Raven, Small and TadmorFergus 1996, 215). Ann Fisher’s 1773 London and Newcastle dictionary was “In Twelves, Price Bound 2s,” as advertised in the York Chronicle; and Weekly Advertiser (Reference FisherFisher 1773b, 381). She marketed it as more modern – distinguishing <i/j> and <u/v> as separate letters and containing more “modern words” (Reference Fisher1773a, iii, v; Reference 738Mugglestone, Johnston and MugglestoneMugglestone 2012, 142). It is advertised in the York newspaper alongside other northern productions (many hers) as “Designed for the use of schools, clerks of offices, or the pocket” (Reference Fisher1773b, 381). Social differences and aspirations are also underscored by the pronouncing dictionaries that began to proliferate around this time as well as by later and cheaper re-editions and manuals. Some of the earliest ones were written by outsiders – by the Scot James Buchanan, the Irishman Thomas Sheridan, and the radical Thomas Spence of Newcastle-upon-Tyne, who hoped to codify “proper” pronunciation in a reformed spelling that would facilitate literacy and resistance among “the laborious part of the people” (quoted by Reference BealBeal 2009b, 159).

22.5 Industrialization

The increasing impact of industrialization in everyday life illustrates some ongoing and broader lexicographical dilemmas. What place do specialist terms have in general dictionaries? What is the definition of a dictionary more generally, and what is its relationship to the encyclopedia? Fisher had claimed that her dictionary contained “a much larger collection of modern words than any Book of the Kind and Price extant” (Reference Fisher1773a, t.p.). But writing in the northeast, Fisher did not register the contemporary lexis of textile industrialization emerging to the west, exemplified below. Nor did the dictionary codify local industrial terminology: hostman, overman, pitman, and sinker pertained to coal mining in nearby Whickham (Reference Levine and WrightsonLevine and Wrightson 1991, passim). Lexicographical practice explains the absence of obviously modern words: Fisher was typical in copying many entries from her predecessors, and these entries were more likely to be for words like mace and waggon (Reference Rodriguez-Álvarez and María EstherRodríguez-Álvarez and Rodríguez-Gil 2006, 310–311). McIntosh finds Samuel Johnson’s Dictionary (1755) more precise and scientific than its predecessors, but it too excludes such specialized and localized terminology (Reference McIntosh2020, 74–79).

Indeed, some obsolete technological and “feudal” words do not appear even in the OED: seventeenth-century dictionaries and the Dictionarium rusticum, urbanicum & botanicum published in 1704 and often attributed to John Worlidge (d.1693) are among McIntosh’s supplementary sources furnishing terms relating to common-field agriculture (Reference McIntosh2020, 14–42 and 126). Were neologisms slow and dictionaries even slower to register concomitant social concepts (Reference CorfieldCorfield 1987, 56–57; Reference CorfieldCorfield 2013)? The phrase working class is used fairly early (1757) by Malachy Postlethwayt (OED3, working class, n. and adj. A. n.), although if Eighteenth-Century Collections Online (ECCO) is accurate, seemingly not in the Universal Dictionary of Trade and Commerce (1751–1755) that Postlethwayt translated from French. Meanwhile as agricultural enclosures inexorably threatened the sense of commoner ‘One who has a joint right in common ground’ (S. Reference JohnsonJohnson 1755, sense 4), rights in common were increasingly enumerated in re-editions of Giles Jacob’s New Law-Dictionary (1729).

In the thirty-nine-volume Cyclopædia; or, Universal Dictionary of Arts, Sciences, and Literature compiled by Abraham Rees “with the assistance of eminent professional gentlemen” (1819, t.p.), the inventions of the “Jenney in 1767,” “Twist, or Water Frame,” and mule (“about the year 1776”) feature in the entry for cotton ([1808] 1819, vol. X). Rees considered his Cyclopædia in the context of the “Scientific Dictionary” (vol. I, vi), defining the dictionary rather broadly (vol. XII). My search of the (imperfectly) digitized ECCO suggests that the earliest dictionary to mention spinning jenny was not a lexical one but the first volume of a New and General Biographical Dictionary, in its entry for ARKWRIGHT (Sir Richard), the perfector of the jenny and the inventor of the water frame (1798, 463). Industrial terminology appeared in cyclopædias like Rees’s, then increasingly in nineteenth-century general dictionaries, especially end-of-century American ones like the Century Dictionary (Reference CmielCmiel 1990, 226; Reference LandauLandau 2009a, 205). So too did other specialist terminology, of course.

Industrialization would affect not only the content but also the physical nature, and thus the audiences, of dictionaries. Rees’s Cyclopædia first appeared serially: hand set, it was printed incrementally from 1802 until 1819 (Cyclopædia 2021). Its additional six volumes of engraved plates, like the engravings and other illustrations in the six-volume Century Dictionary (1889–1891) marked its targeting to the upper end of the market (Reference LandauLandau 2009a, 203–206). But through the nineteenth century technological progress also made unabridged dictionaries more affordable to mass markets, and more portable: “The dictionary spread by shrinking” (Reference CmielCmiel 1990, 82). Stereotype plates had superior “longevity,” “storage” and “mobility” – through space and time – and “enabled a printer to print quickly and relatively inexpensively” – “a small school edition,” for instance (Reference MartinMartin 2019, 108 and 303). In her history of American Dictionaries of the English Language before 1861, Eva Mae Burkett describes Lyman Cobb’s tiny Ladies’ Reticule Companion ([1834] 1841) as “the smallest American dictionary” she had examined – a very portable 818-page reference work for spelling and accentuation (Reference Burkett1979, 70–71). Nineteenth-century dictionaries of engineering reflect the same bifurcation of form, from ambitious (and sometimes abortive) encyclopedias to pocket dictionaries (Reference HoareHoare 2009, 58).

At the same time that dictionaries were being produced and distributed more easily, compulsory education (e.g. with the 1870 Elementary Education Act) was being extended and language standardized in the consolidating nation-state. Mollier and Cachin describe the second half of the nineteenth century as heralding “a second revolution of the book,” with “rotary presses around 1865 and Linotype and Monotype systems from the 1890s” making printing companies like Nelson’s in Edinburgh “industrial manufacturers similar in status to the captains of the textile and steel industries” (Reference 737Mollier, Cachin, Eliot and RoseMollier and Cachin 2007, 305). Nelson dictionaries in the British Library catalog include re-editions of Walker’s Critical Pronouncing Dictionary, a tiny 9cm 510-page Johnson’s Diamond Dictionary of the English Language (1849), a Household English Dictionary (1872), various kinds of Bible dictionaries – illustrated, concise, etc. – and (after a proliferation of colonial primers and readers) the Nelson’s “Highroads” English Dictionary (1914).

New distribution networks brought books to department stores and railway bookstalls (Reference 737Mollier, Cachin, Eliot and RoseMollier and Cachin 2007, 306). And department store mail-order catalogs sent dictionaries across expanding nations and colonies. While the earliest mail order catalog for Canada’s T. Eaton company in 1884 did not include books, the 1894 catalog listed foreign language dictionaries published by Routledge and school dictionaries published by Webster, the latter at three separate price points (Eaton 1894, 175). The idea of a shared history of English was one center for empire. The editors of the OED were slow to adopt technology such as typewriters (Reference MugglestoneMugglestone 2005, 210). But the global public reading program for what was to become the Oxford English Dictionary was indebted to the postal system (Reference OgilvieOgilvie 2013, 67). Editor-in-chief James Murray (like Joseph Wright, his counterpart on the English Dialect Dictionary) had a “rural-industrial background”: the son of a Scottish tailor, as a local schoolmaster he profited from (and provoked) the intellectual stimulation of university vacation courses and local societies (Reference JoyceJoyce 1991, 210).

Compulsory elementary education was not synchronized with universal suffrage. Like books more generally, dictionaries epitomized transitions for some sociopolitical reformers. One pioneer of press freedom and print culture, William Hone (1780–1842), had received John Entick’s dictionary to commemorate his commencement of school, and then Nathan Bailey’s, over which he “incessantly pored” (1912, 44). The illegitimate son of a dyer, Chartist Thomas Cooper (1805–1892) specifically mentions James Barclay’s “quarto” dictionary among the “improved order of books” he had access to when attending the neighborhood day-school (1872, 33–34). After the Mechanics’ Institute movement “began to take shape” in the mid-1820s (Reference Lawson and SilverLawson and Silver 1973, 255), the Society for Diffusion of Useful Knowledge suggested a catalog of books for their libraries that included dictionaries of law, biography, painters and painting, commerce, architecture, mathematics, chemistry, surgery, English (Johnson’s, 1834) (1839, 169–195). For working men, dictionaries seem to have represented knowledge that was “useful and respectable” (1839, 101), while higher society was violently divided about extending the political franchise beyond qualifications of property, and increasingly concerned with distinguishing the “respectable” poor from the residuum (Reference ThompsonThompson 2013, 73–78) as the Second Reform Act (1867) extended the franchise to some working-class men. For wealthier individuals, industrialization fueled consumerism: Eliza Meteyard included a Glossary of Terms in her Wedgwood Handbook: A Manual for Collectors (Reference Meteyard1875, 389–413) (Reference RussellRussell 2018, 91).

Industrialization has a more complicated relationship with domestic service and other forms of work (e.g. Reference BurnetteBurnette 2008). While women’s labor was shifting from agriculture into manufacturing, many still worked in domestic service (Reference BergBerg 1994, 118). Household dictionaries continue to instruct in production of ale through universal cement and yellow dye (e.g. Reference EatonEaton 1823). Interestingly, the first city Directory of Manchester (1772) was published by cookery writer Reference 710CoxElizabeth Raffald (bap. 1733, d. 1781), author of the Experienced English Housekeeper (1769) though perhaps more relevantly also the proprietor of a grocery business, a coffee-house, a cooking school, and a servants’ registry (Reference 710CoxCox 2004). In this directory “every Inhabitant of the least Consequence” (including two male Raffalds) was listed alphabetically by occupation and address – from Achers the silk and linen manufacturers to William Young the Warehouseman (Reference RaffaldRaffald [1772] 1889). By the next year, many addresses had been refined with street names and new house numbers (Reference Raffald[1773] 1889). Proliferating designators of occupations in such metropolitan directories more generally reflect how industrialization led to specialized work (Reference CorfieldCorfield 2012, 34). And these directories exemplify traditional ordering principles intersecting with transforming industrial processes. J. Brown’s The Directory, or List of Principal Traders in London (1732) provides the OED3 with its earliest example of directory ‘an alphabetical list of people or institutions’ (directory, n. 3a; Reference McIntoshMcIntosh 2020, 77).

22.6 Urbanization

Migration to cities highlighted social difference and seemed to accelerate both linguistic leveling and linguistic change. Dialect dictionaries sought to preserve the diverse voices that urbanization juxtaposed, merged, and prospectively silenced. Reviewing the comprehensive Glossary of the Lancashire Dialect (1875) by the city of Manchester’s Literary Club, a journalist in the Salford Chronicle contrasted the “mongrel emasculated lingo” of the city with what he felt was the pure and authentic dialect of the “rural districts” (quoted by Reference HakalaHakala 2010, 74–78). Anxiety about railways and mass education explicitly inspired many leisured codifiers of other rural dialects. Repeating a common refrain (Reference BealBeal 2009a, 140), Margaret Ann Courtney (now known as a folklorist) blamed railways for replacing Cornish speech with “vile Cockney” (Reference Courtney and CouchCourtney and Couch 1880, 1), while before she wrote novels and gardening guides Edith Chamberlain claimed that “the teaching of certificated masters in government schools” was “modif[ying]” the speech of West Worcestershire (Reference 672Chamberlain1882, vii).

Georgina Jackson (a wine merchant’s daughter) claimed to authenticate every word personally, riding in third-class railway carriages on market days to elicit information from blacksmiths, butchers, wheelwrights, farmwives, and the elderly (Reference JacksonJackson 1879, ix–xiv; Reference 749SkeatSkeat 1896, lxviii–lxx). Jackson associated the words she collected with social as well as temporal distance: reading the words tine/tining in Isaac Taylor’s Words and Places (1864) reminded her both of her childhood and of being teased for “speaking like a little Shropshire village-child” (ix). But she takes for granted that “many a word” known by rural “old people” is “now dead or dying out” (xi). One obsolete word denoted “a country weaver” rendered useless by “the factory.” The nog-man is exemplified for Jackson with a quotation from the area of Pulverbatch: “Poor owd Spake [Speake] the nog-man called to beg a spot o’ drink; it’s ’ard times ŏŏth ’im now nobody spins – ’e tells me ’e gets a bit o’ yorn from the factory, an’ waives it ’imself […]” (302). Amateur dialectologists were expected only to compile; the editorial team of the English Dialect Dictionary (EDD) synthesized (1898–1905). Having enumerated “the ways in which large-scale lexicography relied on women as part of their workforce,” Russell uses the case of Joseph Wright’s EDD to observe that these were unchanged by “the move from domestic to institutional space” (Reference Russell2018, 179).

Lexicographers’ commodification of select London languages shows some other outcomes of urban diversity. Certainly the consolidating status of polite London pronunciation can be seen in Benjamin Reference SmartSmart’s 1836 re-edition of Walker’s 1791 pronouncing dictionary. Smart describes the correction of “rustic utterance” as “Londonizing” it, while continuing to condemn the “cockney” (1836, xli), and among his credentials were “having been ‘born and bred at the west end of London’” (xi; quoted by Reference BealBeal 2009b, 173–174). Although slang exists everywhere, the social mixing in cities makes unfamiliar language more prominent. For provincial visitors, fashionable and (more frequently) criminal London slang was commodified in glossaries (see, e.g., Reference ColemanColeman 2004b, 203–205 and 208–210).

The slang shared in Pierce Egan’s Life in London (1821) “caught on throughout the country at all levels of society” (Reference ColemanColeman 2004b, 156). Indeed, Reference EatonEgan’s 1823 re-edition of Francis Grose’s Classical Dictionary of the Vulgar Tongue (1785) surely inspired Dickens’ Oliver Twist (1837) and words like beak ‘magistrate’ and trap ‘police-officer’ (Reference ColemanColeman 2004b, 174). Much slang came from the slums of quickly expanding cities, though Egan (like many lexicographers) draws so extensively on older codifiers that the currency of words is hard to chart. Some terms suggest demographics: Egan inconsistently removed some “Irish” terms from his sources (and added more “Scottish” ones) (Reference ColemanColeman 2004b, 170). Other terms suggest activities: Egan provides the Oxford English Dictionary with some first citations, including to cross ‘cheat’ and fogle-hunter ‘a pickpocketer’ (Reference ColemanColeman 2004b, 173). Egan also provides OED3 with its first (1829) citation of navvy ‘a construction worker,’ shortened from navigator ‘a labourer employed in the construction of (originally) a canal, (now frequently) a road, railway, etc.’ (navvy, n. 1, 1). The effects of urbanization and industrialization on family structures remain much debated, even with the evidence of the census (Reference Puschmann and SolliPuschmann and Solli 2014). According to the 1851 census, railway navvies worked a mean distance of forty-five miles from their place of birth (Reference Gilbert, Southall and DauntonGilbert and Southall 2001, 607). This figure may bridge both local agricultural laborers and peripatetic professional navvies, stereotypically but not necessarily Irish (Reference BrookeBrooke 1989, 39).

Through the nineteenth century, “far-reaching changes […] touched household affairs” “at electric speed,” justifying the publisher’s revision of Isabella Reference 670BeetonBeeton’s 1861 Book of Household Management (1907, vi). The increasing comfort and complication of modern life is reflected by the end-position of the originally initial analytical index and by the increasing use of alphabetization more generally. By the 1907 edition, for instance, the now-alphabetical listings in the much-expanded chapter of “Legal Memoranda” include explanations of rights to light and air (1988–1989) and refer problems relating to “infectious diseases” and “nuisances” and “sanitary requirements” to local government such as the “London County Council” (est. 1889) and to the appropriate “sanitary authority” (1952–1953). Through the nineteenth century, “the scale of investment in the infrastructure of urban services – roads, railways, sewers, water, gas, electricity – was huge” (Reference Daunton and DauntonDaunton 2001, 29). In London, for instance, an intercepting sewage system was constructed in the years after the “Great Stink” of 1858 (Reference Luckin and DauntonLuckin 2001, 212). For this period, OED1 recorded topical sewer-related formations like sewer-ditches (1851, sewer, sb.¹ 5), sewering (1865, sewer, v.¹), pumping stations (1868, pumping, vbl. sb. d.), and ABC process, a process for treating sewage defined at the time as a ‘process in making artificial manure’ (1879).

Reflecting dramatic changes in the capital and capitalizing on the association of a now-dead popular author with his expanding, transforming city, the phrase pumping station appears (though not as a headword) in two London dictionaries compiled and published by Charles Dickens, Jr. Pumping stations feature among Legislation, Projects of from the 1882 edition of Dickens’s Dictionary of London, while the “Abbey Mills Pumping Station, one of the curiosities of modern civilization,” heads the entry for Northern Outfall in Dickens’s Dictionary of the Thames from its Source to the Nore (1880). Dickens, Jr. used the alphabet to define topics ranging from Abstract of Contents to Zoological Society of London (1882): better-established guidebooks like Murray’s Handbook to London as it is (1851) and Baedeker’s London and its Environs (1878) were not called dictionaries (Reference BooneBoone 2006, 178). Proliferating Constituencies, Parliamentary marked the gradual recognition of cities in general electoral reform (Reference DickensDickens 1882). The alphabet was not necessarily the most appropriate organizing mode: despite the ABC in railway timetable titles like Dickens’s Continental ABC Railway Guide, the “tabular” arrangement became “the dominant design” (Reference EsbesterEsbester 2009, 95). But for urban travellers, Dickens’ London dictionary used the alphabet in charts calculating distances between places, rationalizing this challenge of the expanding city.

22.7 Racialization

As British dominions proliferated, dictionaries increasingly depicted cultural and racial differences. Re-editions of Mrs Beeton’s Household Management expanded the original wordlist of French culinary terms to include other languages (1861, 44–46; 1907, 1652–1672), although the prestige of French was preserved in the chapter correlating “English and French names of Articles of Food” (1673–1675). And with Britons increasingly “living under other skies” as well as “welcoming guests from abroad” (vii), chapters were added on “Colonial and Foreign Cookery” representing French, German and Austrian, Italian, Spanish, Jewish, Australian, South African, Indian, American and Canadian Cookery, in that order (1525–1631). In 1883 in Toronto, Mrs Clarke’s Cookery Book also included A Complete Dictionary of Christian Names; their Origin and Signification, a list “prepared at great expense and trouble” (1883, 375). As Patrick Hanks observes (Reference Hanks2009, 141–142), freestanding dictionaries of Christian names were a relatively recent Victorian trend. In a period when women were stereotyped as choosing baby names from novels (Reference HasfjordHasfjord 2016, 14–15), novelist Charlotte M. Yonge’s influential History of Christian Names (1863, 1884) perhaps inspired other women codifiers of What to Name the Baby (Reference ClarkeClarke 1883, t.p.). Mrs. Clarke was far from the old country: her recipes use both (South Asian) “Indian” curry powder (93–98) and (North American) Indian meal (169–171), a phrase Richard W. Bailey uses to exemplify settlers’ “reluctance to confront the new linguistic environment” by adapting familiar English words (1991, 73–74). The babies’ names Clarke codified show the importance of other distant traditions, with etymologies including Hebrew, Greek, Latin, “Old. Ger.,” “Ang-Sax.,” and Celt (375–385). The role of history in the rise of this imperial nation state can be exemplified not only by the OED but also by the revival of Anglo-Saxon names like Alfred and Edith (“Old Ger.” in Clarke) in the nineteenth century (Reference HanksHanks 2009, 139).

As British colonizers encountered human difference, dictionaries of Indigenous languages reflected generally unarticulated motives to exploit or assimilate. One early wordlist compiled in what is now Canada was obtained from Oubee, a young Beothuk girl who had been captured in 1791 and later brought to England, where she died (Reference ConsidineConsidine 2003, 251). Not all North American linguistic informants were Indigenous, women, or kidnapped (Reference BaileyBailey 1991, 64–69). But women were vulnerable to “transculturation” (Reference Thomas, Ayres-Bennett and SansonThomas 2020, 332), and three Indigenous women furnished the earliest and indeed the only wordlists of the extinct language Beothuk: Oubee, Demasduit (c. 1796–1820), and Shanawdithit (c. 1801–1829) (Reference HewsonHewson 1978, i). Oubee’s vocabulary was apparently elicited in an effort to end hostilities (Reference MarshallMarshall 1996, 429), but despite the later codification of phrases meaning “don’t be afraid,” “not hurt you” and “we come to be friends” (Reference MarshallMarshall 1996, 214), by Shanawdithit’s time Beothuk numbers had dwindled. European disease was one factor: both Demasduit and Shanawdithit died of tuberculosis. British exploitation and settlement of their coastal territory was another: the Beothuk were malnourished inland or murdered by fishers and furriers, sometimes in revenge for theft. With Shanawdithit’s death in 1829, the Beothuk people and their language became extinct (Reference Pastore and StoryPastore and Story [1987] 2003). The poor documentation of their language reflects British focus on natural resources.

In North America, wordlists embedded in the records of “explorers” like Alexander Mackenzie reflect the linguistic diversity of the area while representing English as the common denominator for what he calls the Knisteneaux and Algonquin languages (1801, cvii–cxvi), Chepewyan (cxxix–cxxxii), and Nagailer and Atnah (257–258) and embedding such cultural knowledge in the implicitly economic “General History of the Fur Trade from Canada to the North-West” (i–cxxxii). And although many missionaries codified Indigenous languages through the nineteenth century (Reference Goddard and GoddardGoddard 1996), religious assimilation heralded linguistic and cultural assimilation. Some later vocabularies of Pacific Indigenous languages were collected by US government-sponsored surveyors as well as by traders including representatives of the Hudson Bay Company (Reference Goddard and GoddardGoddard 1996, 36–39). Reflecting these contacts, settler Englishes and their dictionaries (including and imitating the OED) contain mostly borrowed nouns denoting flora and fauna, places and people – despite some editorial debates about these words’ status as English (Reference 702BensonBenson 2001, chapter 6; Reference OgilvieOgilvie 2013, 39; Reference Rice, Bergs and BrintonRice 2012, 1760–1763).

Social dealings and distinctions between Europeans and others were increased by railway labor and by the settlement it accelerated. Stereotypical conceptions of human difference were articulated in John Sandilands’ Western Canadian Dictionary and Phrasebook (Reference Sandilands1912, Reference Sandilands1913), as the Canadian Government encouraged and justified the West’s settlement and “civilization” by British immigrants (Reference DohertyDoherty 2020, iii–iv and 9). Himself an immigrant to Western Canada, the Scot Sandilands used many racist terms to denote other inhabitants of Canada (Reference DohertyDoherty 2020, 9 and 15). In the entry for zebra ‘a person of undiscoverable nationality,’ phrases including the words white, red, and yellow show us that skin color was the criterion that “divides mankind up” for “[t]he haphazard student of ethnology in the Western camps” (zebra, 1913). According to Nicholas Hudson, the biological classification of humans arose in part from the “new science of human taxonomy” devised by Enlightenment thinkers like Buffon and Linnaeus. The word race, for instance, “gradually mutated from its original sense of a people or single nation, linked by origin, to its later sense of a biological subdivision of the human species” (Reference Hudson1996, 258). Contemporary intellectual conceptions of race were registered belatedly in mainstream dictionaries. Hudson finds the now-current meaning of race appearing in the 1835 edition of the Dictionnaire de l’Académie française but not in any nineteenth-century edition of Webster’s or Chambers’ dictionaries. In 1910, the eighth volume of the first edition of the Oxford English Dictionary “authorize[s] our familiar modern sense of this word: ‘A group of several tribes or peoples, forming a distinct ethnic stock’ and ‘One of the great divisions of mankind, having certain physical peculiarities in common’” (Hudson 247, 259n3). Hudson explains these “generalized [cultural] impressions of whole continental populations” as a consequence of European economic exploitation (258).

(How) do lexicographers represent their society’s racial prejudices? According to Coleman, the earliest cant and slang dictionaries claimed to codify the language of the “dispossessed poor” and of the Roma specifically “as a defensive tool for the magistrate and the householder.” Romany words were a selling point, though few were actually codified (Reference ColemanColeman 2004a, 2 and 7 and 22). Some slang made links and distinctions between Britons and others: Janet Sorensen argues that the novelist Defoe transformed terms like B.E.’s definition of the Black-Indies as ‘Newcastle, from whence the Coals are brought’ (1699) ultimately to distinguish British laborers as only figuratively Black, in order to distinguish a true (white) British liberty (Reference Sorensen2017, 79–81). Yet racial prejudices are also reflected by absence in dictionaries. Kenneth Cmiel claims that, in North America, racial relations became extensively represented only in the encyclopedic dictionaries published near the end of the nineteenth century – the Century Dictionary (1889–1891); the 1890 Webster’s International; Funk & Wagnalls Standard (1893–1895); and the American Encyclopædic (1894). Cmiel sees the size of these dictionaries reflecting the progress and intersection of philology and science, broadly speaking (Reference Cmiel1990, 223–232). But he observes that the increasing tolerance of slang and jargon led to a shift not just to “informality” but to “insult[]”: the new dictionaries added or sanctioned more “ethnic and racial slurs” insulting African-American, Southern European, Jewish, and Mexican people. While some terms had usage labels like “vulgar” or “opprobrious,” others might simply be flagged as “low,” “slang,” or not marked at all (Reference Cmiel1990, 231). Even the existence of labels often confirms a racist perspective. For instance, Doherty reports that the only racist term in the 1912 edition of the Western Canadian Dictionary “actually labelled as derogatory” “was considered derogatory […] only when used to describe white people.” Siwash meant both ‘a native Indian’ and ‘a mean, contemptible, worthless, person; a term of address that is meant to give offence’ (quoted by Reference DohertyDoherty 2020, 10). Phil Benson concluded from “warning labels indicating that the usage is offensive or potentially offensive” that “the cultural centre of the British and US dictionary […] is essentially that of the Anglo-American white male” (Reference 702Benson2001, 50–51). The treatment of racial epithets along with other offensive terms remains a challenge in descriptive lexicography.

22.8 Concluding Summary

Dictionaries were devised to aid communication among speakers, so their content both negotiates and represents social differences. In diglossic societies ranging from medieval Christendom to settler colonies, monolingual dictionaries codifying prestige languages might not represent vernaculars at all. If vernaculars rise in status, the content of wordlists might correlate with and be prompted by social factors – farming and carving terms for groups new to noble estates; Latinisms for ladies, perhaps especially maternal or Puritan. Some wordlists may be sequestered in separate texts – rural words in dialect glossaries, or the terminology of textile manufacturing in encyclopedias. Indeed, some wordlists appear in other genres entirely. Wordlists embedded in centuries of household manuals suggest the symbolism of archaic carving terms and the rising importance of addressing and spelling letters correctly, as well as the kinds of households that become newly visible to the increasingly literate. New alphabetically organized genres order the names of business owners and institutions, as increasingly complex cities perplex residents and attract newcomers. The structure of wordlists may also reflect social factors – rhyming vocabularies for mothers’ oral instruction; alphabetical indexes of recipes for labor-linked literacy; tables equating colonial English with many Indigenous languages, embedded in economically motivated exploration accounts.

The production and distribution of dictionaries can index shifting social categorizations. Women’s domestic roles can be exploited and extended in attributions that are feminine but pseudonymous – the appropriation of the name Hannah Wolley by a now unknown compiler of household manuals. In a country without a language academy, multiple agents trace boundaries of the language and the nation – or empire. Title pages, dedications, and advertisements might align texts with audiences. Developing industrial processes allow the “same” text to be produced for different demographics and settings: as spelling and pronunciation become standardized and primary education spreads, portable texts might be needed by more people. Catalogs quantify hierarchies among dictionaries – and department store catalogs reveal the dependence of Canada’s small market on publications from Britain and the United States.

How can one measure the impact of dictionaries? Booksellers’ records are partial (in both senses of that word): sale catalogs of individual libraries were often padded, and (like ownership records) sales ledgers rarely survive. The audiences invoked by authors (or booksellers) were not always the users of dictionaries. And the linguistic information in dictionaries is not always immediately influential: contemporary corpus-based studies can confirm that some of Noah Webster’s recommended spellings took many decades to spread, despite their patriotism (Reference Mair, Yáñez-Bouza, Moore, van Bergen and HollmannMair 2019). The innovation of dictionaries indicates their perceived need for their intended readership. And the revision and the plagiarism of dictionaries confirms another kind of influence. Various social groups used these records to their advantage. And new dictionaries also illuminated aspects of society which might not have been visible before – serving as instruments of ideology as well as repositories of information.

23.1 Introduction

Philology and linguistics emerged as western scholarly disciplines centuries after the first European dictionaries appeared, and over time they helped improve dictionaries in various ways. In turn, dictionaries supplied philology and linguistics with essential evidence, especially as they influenced lexicographical methods according to their own interests. Major English and American dictionaries of the nineteenth and twentieth centuries, for instance, had philologists among their editors, sometimes brought onto the masthead to write etymologies, if nothing else. The historical relationships among dictionaries, linguistics, and philology are clear, but they are beset by ironies or paradox, depending on one’s point of view.

One would think, for instance, that having philologists and, later, linguists as editors, dictionaries would treat philological and linguistic terms well, but they do not. Even the terms linguist and philologist are less adeptly defined than one might expect. Lexicography thus fails to serve its disciplinary allies, even when those allies are writing the dictionaries. The observation seems trivial at first sight, for no dictionary can include all words. In fact, it is non-trivial: it shows how dictionaries implicitly register the cultural value of knowledge. At present, computer terms are better represented and more effectively defined than terms about language and the study of language.

By their very structure, dictionaries are insufficient vehicles of language history, an observation that may strike one as counterintuitive. A dictionary entry must abstract information – forms, meanings, regionality, social level, origins – as though there were no question about the conclusions they report. But language history is so complex and messy that it fares better in encyclopedias than in dictionaries: the more encyclopedic the dictionary, the more likely it can do justice to philological knowledge about a language. The relative status of encyclopedia and dictionary is typologically problematic (see Hancher 2019), and nothing shows this more clearly than the attempt to convey philological knowledge in narrowly conceived dictionary types. As regards philology and linguistics, we are best served by a variety of dictionary types. Unexpectedly, philology and linguistics are case studies in the limits of lexicography, as traditionally understood.

To grasp the ironies, however, one must start with the history.

23.2 The Concepts of Linguistics and Philology: A Historical View

I remember my surprise when I discovered that the Encyclopedia Britannica, at least beginning with the glorious eleventh edition, for some time, had numerous entries containing mostly or even only the etymology of the word at the head. Here are a few random examples.

The entry almond begins with a detailed etymology (Latin, Greek, Arabic, and Italian) and a comparison of the word’s form in English and French. At goose, the gray lag goose is mentioned, and the reader will find a nontrivial derivation of lag: not only lag = laggard but a possibly sound-imitative base. The entry heaven is truly encyclopedic: etymology, usage, the number of heavens in different beliefs, and the words for ‘heaven’ in other languages. One can go on and on.

Those who deal with technical terms know that frequently the sought-for definition turns up in an encyclopedia, rather than in a dictionary; hence the existence of “hybrids.” Two such products stand on my shelf: one in Russian, Entsiklopedicheskii slovar’ (1953–1955), a depressing monument to the darkest years of Stalin’s dictatorship; the other in English, The Century Dictionary: An Encyclopedic Lexicon of the English Language (1889–1911), a splendid piece of scholarship, a true blend of both genres: a heavily illustrated, dependable dictionary and a store of up-to-date information.

In what follows, more hybrids will emerge, linguistics and philology among them. Even as long as we remain in the field of lexicography proper, some genres tend to merge. For instance, the German tradition has established two types of dictionaries: etymological and historical. The first goes back to the early 1600s; the second is the brainchild of the Neogrammarians (Junggrammatiker), a school that dominated German and partly the rest of European linguistics in the second half of the nineteenth century and to which we owe basic textbooks with titles like Old English Grammar. In theory, the etymological dictionary was supposed to discuss the origin of words, that is, go into the prehistory of the recorded forms, with the historical dictionary picking up where the etymological one left off. The author of the first Neogrammarian etymological dictionary of German was Friedrich Reference Kluge and MitzkaKluge (1883; by today’s standards, a modest but highly successful venture). The production of the German historical dictionary was entrusted to Hermann Paul (1897), the head of the Neogrammarian school.

The material in the two dictionaries partly overlapped. Kluge and his successors could not establish a word’s origins without referring to the entire spectrum of its later meanings and even without a look at its development in the Standard and in dialects (for, otherwise, the modern speaker often would not have been able to understand how the story began), while Paul had to touch on the form and meaning of the oldest recorded words to the extent that they continued into the modern language, though, naturally, not in such detail as was expected of Kluge.

Not unexpectedly, both lexicographers encountered familiar difficulties, which have not always been overcome by later editors. For example, in Alfred Schirmer’s excellent sixth edition of Paul’s dictionary (1959), we read (in my translation from German): “Borax ‘borsaures Natrium,’ late Middle High German buras, taken over from Persian būräh via Arabic.” Schirmer’s contemporary Walther Mitzka, the editor of the twentieth edition of Kluge, had only this to say about Borax: “Persian būräh ‘borsaures Natiron (sic)’ reached European languages from Arabic būraḳ, baurag. Late Middle High German buras, early New High German borros go back to Middle Latin. This is the source of Modern German Bor, Borsäure, Borwasser.” The difference between the two entries is not great. Elmar Seebold, Kluge’s latest editor, wrote a long entry on borax. He was more acutely aware of the familiar question “dictionary or encyclopedia?” and went into detail about the travels of the term from language to language and the changes of its meaning over the centuries. But, naturally, in dealing with such words as sitzen ‘to sit’ or sieben ‘seven,’ in which the referent needs no explanation, the dictionaries go different ways. The question about the extent to which dictionaries deal with words versus things brings us to the difference between linguistics (clearly a study of words, their use and origin) and philology (a study broader and vaguer, to which we return below).

The philology-versus-linguistics opposition did not exist for Paul, Kluge, and their contemporaries abroad (just as it did not exist for Rasmus Rask and the Grimm brothers), because none of them had any difficulty in moving from period to period: in their day, it was taken for granted that specialists we today call linguists had full command of the history of language and, as a matter of course, had learned Latin and Greek at school. In the English-speaking world, linguistics was a rare word until roughly a hundred years ago. Linguistique enjoyed greater popularity in France. The Germans used the all-embracing term Sprachwissenschaft ‘language science,’ with its broad connotations.

In both Europe and the United States, the study of language in college and university often neglected the achievements of contemporary scholarship and meant more Latin and Greek. All the prestige went to those who taught them – for what, one wondered, is there in the modern languages that deserves serious attention? Writing English, French, and German well, along with oratorial brilliance, also came as the result of following the classical models. Strangely, this medieval attitude toward language did not interfere with the universal admiration for writers like Charles Dickens, who knew neither Latin nor Greek. The spread of linguistic knowledge, divorced from the Classics, depended on the extent the “newfangled” theories made headway in and outside Germany. England proved especially conservative. The idea of an English department would have seemed ridiculous even to many contemporaries of Henry Sweet (1845–1912) and Joseph Wright (1855–1930), and almost all research into Old and Middle English was done by Germans.

Attitudes began to change rapidly shortly before World War I, mainly under the influence of Saussure’s Cours de linguistique générale (Reference Saussure1916). Though not without resistance, structuralists and structuralism won the day in Europe (N. S. Trubetzkoy and Roman Jakobson; later, Louis Hjelmslev and John Rupert Firth) and the United States (Edward Sapir and Leonard Bloomfield) – all that despite the numerous disagreements among the structuralists, especially acute in American mid-century scholarship. A barrier was set up between synchronic linguistics and philology; the latter concentrated on diachrony. The previously unthinkable concept of modern philology came into being, an area of study that did not draw its inspiration from Saussure, yet its object was the history of the post-Classical languages. Though Germany remained untouched by the structuralist wave, it continued to lead the way in the study of the older periods of the Germanic languages. In the field of Romance, it also dominated, but it had outstanding competitors in Switzerland and France. Here, only a few famous names will be mentioned: Gaston Paris (perhaps the most influential scholar in this area), Antoine Thomas, and Thomas’ formidable opponent Hugo Schuchardt (they differed in their attitude toward the role of semantics versus phonetic laws in etymology; see a few remarks on this subject in Reference LibermanLiberman 2009, 217).

The Saussurean revolution had an unexpected and unwanted side effect. The study of the Classical languages Greek and Latin fell into decay, and after World War II the history of English, German, and other modern languages also followed suit. Linguistics, confined to synchrony, divorced itself from philology (and thereby from a study of literary texts), so that a modern linguist need not have seen a line of an old chronicle or poem in any language. Linguistics became synonymous with language theory. It concentrated on phonetics (but more on phonology) and grammar. Lexicology and lexicography found themselves in no-man’s-land. Saussure, who was also one of the greatest historical linguists of all times, could not anticipate such a result of his teachings.

The digression above is of course common knowledge, but without it the treatment of linguistics and philology in dictionaries might be less clear. At present, linguistics is a broad term, referring to the branch of knowledge that deals with language or, to put it differently, the science of language, which returns us to the German Sprachwissenschaft. However, the implication that the object of linguistics is language theory remains, and many linguists are fond of calling themselves theoreticians. Someone who deals with a crux in Homer or, for that matter, in Shakespeare, or prepares a new critical edition of Beowulf is not a linguist but a philologist, unlike the scholar pondering the moraic structure of long vowels in Greek or the use of the present perfect in Old Icelandic.

In the nineteenth century, the science of language was occasionally called linguistic in English, without the plural ending (like rhetoric and arithmetic), and as late as 1924 the Linguistic Society of America was founded, and Leonard Bloomfield wrote an article titled “Why a Linguistic Society?” The earliest attestation of linguistics in the OED is dated to 1837. A curiosity of English usage is the word linguist, defined in the original edition of the OED as “[o]ne who is skilled in the use of languages; one who is master of other tongues beside his own” (so first in Shakespeare, 1591, and later, down to 1867); (obsolete [!]) “a student of language; a philologist” (with citations from the years 1641, 1695, 1748), and two more obsolete senses: “[a]n interpreter” and “[o]ne who uses his tongue freely or knows how to talk; a master of languages” (1588, 1599, 1612, 1691). In American English, too, only professional linguists call themselves “linguist.” In the wide world outside linguistics, a linguist is a person who speaks several languages fluently – a polyglot. Henry Cecil Wyld, an outstanding British historical linguist and lexicographer, probably called himself a philologist. In his Universal Dictionary (1932), linguist is defined only as “one possessing knowledge of languages; generally of one specially skilled in use of foreign tongues,” followed by the examples “he’s quite a linguist; a good linguist; a poor, bad, linguist, one who has a poor command of foreign languages.” This is all the more surprising as in the much earlier Century Dictionary (1889–1891), the second sense is “a student of language; a philologist.”

Philology, as noted above, concentrates on language history, be it classical, modern, or sacred (the latter deals with the scriptures in Hebrew and Greek). Against the background of this general information, it is instructive to examine the treatment of philology in several dictionaries. But before doing this, we need to remember the meaning of criticism, the word regularly occurring in connection with philology. Criticism, according to Murray’s OED, is “The art of estimating the qualities and character of literary or artistic work […]; [t]he critical science which deals with the text, character, composition and origin of literary documents.” Yet, as we will see, there may be more to it.

Samuel Reference JohnsonJohnson’s (1755) definition of philology includes only two concepts: criticism and grammatical learning. But a century later, John Ogilvie in his Imperial Dictionary (1850) devoted a detailed paragraph to philology, and it deserves to be reproduced in full because it is unusually detailed and because this edition is now rarely consulted, overshadowed by Charles Annandale’s improved version (1882–1887):

Both Johnson and Ogilvie seem to have used criticism as a synonym of analysis. Ideology is defined in Ogilvie as “the science of ideas or mind.”

The entry quoted above is remarkable. It contains the most detailed definition of philology one can find in any European dictionary: philology emerges from the entry as a branch of knowledge devoted to the study of everything in language and literature. We also notice that Ogilvie did not miss the progress of comparative linguistics with its goal of reconstructing the Proto-Indo-European language (Indogermanisch, Ursprache, as it was called in Germany) and discovering the home (Urheimat) of its speakers. His awareness of linguistic work on the Continent is not a trivial fact in English scholarship in the middle of the nineteenth century. That definition of philology will look even more impressive when we compare it with its source and sequel. Ogilvie, as he explained, based his dictionary on Webster. Reference WebsterWebster’s (1828) definition is quite different, however. He, too, mentioned “a love of words” and the rest, but this was a mere gloss on the Greek etymon of philology. Point 2 in Webster’s entry sounds so:

Annandale succeeded in editing the Imperial without destroying Ogilvie’s format and style. When he decided that he could bring out a new version of that splendid work, he did so (Reference Ogilvie and AnnandaleOgilvie and Annandale 1882).

One of the most noticeable differences between Imperial1 (1850) and Imperial2 (1892) is the treatment of etymology. In 1852, when Ogilvie did what little he could to update Webster’s fanciful derivations, no reliable contemporary source existed. We will return to the treatment of etymology in dictionaries; here only a few basic facts should be mentioned. As far as the field of Germanic is concerned, etymological dictionaries do not predate the early seventeenth century (for details see Reference LibermanLiberman 1998 and Reference Liberman1999). When they do appear, we find in them a mix of reasonable conjectures and hopeless attempts to derive old words from Hebrew, Greek, or Latin, depending on the lexicographer’s predilections. Webster ruined his etymologies because, true to his religious beliefs, he divided all languages into Hamitic and Japhetic. Moreover, throughout his life he remained a follower of Horne Tooke and shared Tooke’s bizarre idea that, whenever possible, all words should be traced to verbs. The literature on Webster’s philological methods needn’t be surveyed here, but I would cite Reference 730LairdLaird (1946), not made obsolete by later research (see also Reference FineganFinegan 2020a.)

Webster’s etymologies are no longer interesting from any point of view (who indeed will now discuss the idea that heat, hate, and behest, to give a typical example, share the same root, or refute the derivation of English words from Hebrew?), but Webster treasured his explanations in historical semantics even more than any other aspect of linguistic reconstruction. As is well-known, historical semantics never developed laws like the Neogrammarian laws of phonetics, even though quite a few regularities have been discovered; in this area, clever guesses and reasonable but unsystematized observations still predominate. For example, it remains a matter of debate how temple “shrine” and temple “flat part of the side of the head” are related, if at all. Several hypotheses compete. This is what Webster wrote in the introduction to his dictionary:

This explanation may not convince anyone, but, like all Webster’s explanations of this type, it can still be discussed today, while his etymologies cannot. When it came to etymology, Webster’s dictionary failed his readers.

Between the years 1879 and 1882, the first edition of Skeat’s etymological dictionary was being published in instalments (see Reference SkeatSkeat 1882), and Reference Ogilvie and AnnandaleOgilvie and Annandale (1882) followed him. By that time as well, many of Webster’s explanations had been tacitly rewritten by his later editors, but following Skeat was a major step forward. Also, by 1882, comparative studies had stopped being an exotic novelty in England, and it became possible to give a more specific definition of philology. Annandale first quotes Johnson’s definition of philology and then continues:

A long quotation from Max Müller follows, to prove the point.

As one can see, the definition of philology vacillated between the study of language in the broadest sense of this word, including comparative studies, and what we now call linguistics. In the first half of the nineteenth century, philology was considered to be a branch of literature. For a long time, lexicographers could not come up with a precise definition because the object of philology, besides being “a love of words,” of necessity remained all-encompassing and, consequently, vague. As time went on, the vagueness did not diminish. The most respectable “thick” dictionaries referred to the study of the culture of civilized peoples as revealed chiefly in their languages, literature, and religion, mythology, and folklore; stressed the importance of written documents to the exclusion of oral tradition; emphasized the study of the nature and development of language, and the like, but almost always with the accent on history, except when philology was equated with linguistics.

In our attempts to decide what “a love of words” means to a scholar, we have hardly made much progress since the days of Ancient Greece and Rome. The Century offers an instructive array of definitions (s.v. philology): “love of learning and literature” (Cicero), “explanation and interpretation of writings” (Seneca), (with reference to Greek philología) “love of dialectic or argument” (Plato), “love of learning and literature and history” (Plutarch, etc.). The entry ends so: “The word is sometimes used more especially of the study of literary and other records, as distinguished from that of language, which is called linguistics; often, on the other hand, of the study of language or languages.” The explanation is supported by quotations from William Dwight Whitney (the Century’s editor)Footnote ¹ and again Max Müller. The OED gives examples of the use of philology from 1522 to 2004.

To sum up, linguistics, when understood as language theory, is not philology, while philology, with its interest in language history, old literature, and oral tradition, is in some respects broader and in others narrower than linguistics. In what follows, no distinction will be made between those two areas of knowledge, except to consider the definition of philology in a few modern dictionaries of linguistic terms.

23.3 Some Special Terminology in Dictionaries

While examining the role of linguistics and philology in dictionaries, we should think not only of the lexicographer but also of those who consult the product. A hundred years ago, dictionary users could be expected to know a good deal about grammar and perhaps have a smattering of some foreign language, for the very fact of turning to a dictionary testified to a certain level of sophistication. Today, such an expectation would be unrealistic. A curious example of our surrender to public ignorance will be found in the otherwise excellent German Duden. German, like Latin, has three genders, and the traditional abbreviations for them have for centuries been m. (masculine), f. (feminine), and n. (neuter). Apparently, the Duden editorial board decided that it could no longer rely on the knowledge of such a fact. Every noun is now supplied with the definite article characterizing the gender: der, die, or das. However, the abbreviations for strong and weak verbs (st. V. = starkes Verb and sw. V. = schwaches Verb) have survived. Probably, there was nothing to replace them with. Thus, some grammatical knowledge that earlier generations relied on is no longer available even in a dictionary as distinguished as the Duden.

In the English-speaking world, verbs are classified into regular and irregular, which is fine for teaching purposes but not for “philology.” Even etymological dictionaries no longer cite Greek in its traditional form. All words are transliterated. The same holds for Hebrew. Two hundred years ago, Noah Webster did not doubt that his best-educated readers would be familiar with both alphabets, but his dictionary was an immediate success on both sides of the Atlantic, and even then, a reading knowledge of Hebrew and Greek among the users of his dictionary could not be taken for granted. Once again we see that philological knowledge is discounted even in specialist dictionaries.

Lexicographers are expected to satisfy everyone who may open their product. Consider the unusual title of Daniel Jones’ masterpiece, Everyman’s English Pronouncing Dictionary, as it was commonly called. It will be remembered that the publisher of that dictionary was Joseph Dent, the founder of the famous Everyman’s Library, the series in which the dictionary was first published in 1917 under the title An English Pronouncing Dictionary. But Everyone, addressed in the Library’s motto (“I will go with thee, and be thy guide, / In thy most need to go by thy side”) is hard to define and even harder to satisfy.

A passage from the section “Acknowledgements” in the second edition of The Concise Oxford Dictionary bears repetition: “The first letter we received after C.O.D. appeared was a demand for repayment of the book’s cost, on the ground that it failed to give gal(l)iot, to settle the spelling of which it had been bought.” Reviewers and correspondents regularly comment on the deficiencies of dictionaries, but the user’s level of expertise has hardly been discussed in detail. It is only understood that publishers try to make their product accessible to as many people as possible and have particular groups in view: children, beginners, advanced learners, and “everyone.” Below, only Everyone’s potential attitude toward the terminology of linguistics will be touched upon. (For the others, see Rennie, Chapter 7, this volume, and Hargraves, Chapter 4, this volume.)

The past bequeathed to us a vast vocabulary of grammar (understood in its broadest sense) and rhetoric, from amphibrach to zeugma. How much of it deserves inclusion in a “thick” dictionary? The letters D and E yield unusually many words relevant to Greek studies. I made up a list of words I knew from my experience (I am not a Classical scholar!) and left distaxia, episememe, and their likes out of consideration. Predictably, dactyl, digraph, diphthong, ellipsis, and eponym turned up everywhere, while the Century featured even diptote. Also anaphora, in contrast to epiphora, did very well. To my surprise, until its fourth edition (2000) the American Heritage Dictionary missed ergative, emic ~ etic, and enlargement (its special sense). Since online reference works need not count every inch, they may fill their pages with the most exotic words and provide etymologies, but even they have to draw the line somewhere.

Against this background, the question arises whether anyone ever looks up even digraph or diphthong in a dictionary. (Today, when most people search for, rather than look up, everything they need, the situation is different, but mine is a historical view of dictionaries.) I think it is safe to assume that the student who has come across digraph, diphthong, and their likes could not have stumbled on them in a poem, newspaper article, popular magazine, or novel. Most probably, the word occurred in a textbook in which such technical terms are explained at length or in a work written for the initiated, who have mastered this vocabulary. If deixis and enantiosemy did not make it into our most authoritative dictionaries, why did diphthong? It is indeed a frequent word in phonetics (often mispronounced as “dipthong”) but nowhere else! To be sure, similar problems confront dictionary makers with regard to all special terms, be they in philology, botany, or medicine. The reasons for inclusion are frequency and “general recognition,” but still: does anyone consult a dictionary for the meaning of diphthong?

The old question arises here: where does the dictionary end and the encyclopedia (or even the textbook) begin? The treatment of such a grammatical term as ablative is a case in point. OED offers an exhaustive treatise on the ablative case. The Century, though less expansive, still mentions the situation in Latin, Sanskrit, and Zend, assuming that such references make sense to the user. More recent dictionaries say considerably less, but in the same vein. They follow the formulation of the Shorter Oxford English Dictionary (2002), which chose to write the following about the ablative: “Grammar. Designating, being in, or pertaining to a case in Latin and other languages expressing direction from a place, or time, and variously also the source, cause, instrument and agent, manner, and sometimes place and time of an action or event.” To this a section on ablative absolute is added. It is hard to imagine anyone ever opening a general-purpose dictionary to discover the meaning of ablativus absolutus.

Wyld (Universal 1932) must have realized that a dictionary is not a textbook and wrote this about the ablative: “Name invented by Julius Caesar for a Lat. case in nouns, expressing origin, agent, instrument &c.” Yet he added: “a case with similar functions is found in all the older Aryan [= Indo-European] languages.” Wisely, I believe, he tended to stress the lexicographic aspect of the concept but, unfortunately, could not always decide where to stop. Thus, ablaut appears in his dictionary with this formulation:

If ablaut is the same as gradation, one cannot help wondering why two dissimilar entries were needed.

And a last example, for comparison:

Mutation also found its way into the Universal. Instead of saying something like “the same as umlaut” or “see umlaut,” Wyld anticipated the definition that appears at the letter U and gave a different illustrative example:

One could have dismissed the later entry as redundant, but I wonder what Wyld meant by a prior modification of an intervening consonant (the detail not mentioned under umlaut!). According to one school of thought, i-umlaut caused the palatalization of the postvocalic consonant. Is that what he meant? Incidentally, where the older dictionaries give the marker philology to ablaut and umlaut, the American Heritage Dictionary (1969) says linguistics, the Random House Dictionary vacillates between grammar and linguistics, while the Shorter Oxford, NOAD2, and recent editions of the American Heritage Dictionary (2011) say nothing.

Assuming that words like ablaut and umlaut deserve inclusion in dictionaries, the information should probably be reduced to a bare minimum. Even Hornby’s OALD features ablaut and umlaut. I am quoting from the third edition: “Ablaut: (phil) systematic vowel changes in verb forms of Indo-European languages (as in drive, drove, driven).” To be sure, ablaut permeates the entire system of word formation and is not limited to verb paradigms, but this information lies beyond the scope of explanatory dictionaries, even when they call themselves universal. I find the entry acceptable, but let us read on: “Umlaut: (in Germanic languages) vowel change shown by two dots over the vowel (as in German plurals Männer, of Mann, and Füsse, of Fuss).”

Although the same approach informs both entries, the second one is considerably less successful than the first, and it highlights the difficulty in producing a “popular” definition of such a many-sided concept as umlaut. Wyld relied or pretended to rely on his readers’ awareness not only of Sanskrit but even of Zend. However, someone who consults Hornby’s learner’s dictionary may not have seen a word of German or heard about the existence of Germanic languages, so that the examples (left untranslated; and why ä and ü, but not ö?) added nothing to the definition, perhaps even complicated it.

In an English dictionary for learners from Latin America to China, all one needs is “Umlaut. In some languages, two dots over a vowel letter that change the pronunciation of a, o, and u, among others.” Amusingly, reference to the graphic device (two dots) seemed so trivial to the compilers of the greatest dictionaries that they missed it. The origin and function of umlaut belong to books on historical linguistics. To conclude: with regard to philological (linguistic) terms, dictionaries, as I believe, should not try to compete with encyclopedias and should exercise great reserve in writing definitions.

Even specialized dictionaries may have problems with more exotic terms. Etymological dictionaries presuppose especially well-prepared users. Yet one should count not only on professors of comparative linguistics but also on ambitious beginners. Kluge’s German etymological dictionary has run to twenty-five editions. The last five (1989–2011), by Elmar Seebold, have a section titled: “Introduction to the Terminology” (translated from German) and an index containing close to 200 terms. It would be hard to imagine any user of that dictionary who does not understand the meaning of adverb, infinitive, interjection, metaphor, and neologism. But bahuvrihi, durative, infix, laryngeal, vrddhi, and a few others may puzzle even an advanced graduate student.

We are again reminded of our contemporaries’ inadequate exposure to grammar. Anyone who, at least in the United States, teaches a foreign language at university level, be it Russian or Old English, has to begin by explaining to students the meaning of such elementary concepts as subjunctive, imperative, and passive, let alone declension versus conjugation, suppletive, and their likes. Gone is the time when a dictionary user could read Greek and Hebrew words in their original scripts, had heard about Zend, and knew the difference between the present and the past participle. We are even afraid of the term auxiliary and prefer helping verb. On the other hand, bahuvrihi and vrddhi need glosses at any epoch, but their place is not in all-purpose dictionaries. Once again, then, dictionaries underserve philological inquiry and reflect a decline and perhaps interest in grammatical knowledge.

23.4 Two Linguistic/Philological Components in Dictionaries: Phonetics and Etymology

Every dictionary contains a whole world. Most people use dictionaries to find out what a certain word means and how it is used, spelled, or pronounced. Any debate about whether dictionaries should be descriptive or prescriptive interests chiefly journalists. The public views dictionaries as authorities. The first edition of The American Heritage Dictionary of the English Language (1969; AHD) opens with a series of short essays. Morris Bishop, who wrote about usage, noted quite correctly that, when in doubt, all of us say the same, namely, “Look it up in the dictionary.”

By definition, the dictionary is a sophisticated linguistic-philological product. The broader its scope, the richer its linguistic component. I mention without discussion two obvious things: (1) definitions are subtle exercises in semantics and (2) the division of the headword into senses from the most common to rare (archaic, obsolete, obsolescent) and distinguishing between polysemy and homonymy cross the line between synchrony and diachrony; thus, in a way, all sizable dictionaries are historical, just as, of necessity, all of them are encyclopedic, even if to varying degrees.

The rest depends on the aspirations of the author and the publisher. Dictionaries can highlight or ignore the difference between the main varieties of the language (British versus American English, German in Germany and in Switzerland, French in France and in Canada or Belgium, and so forth), note the extent of phonetic variation, discuss synonyms, indicate the word’s origin or give detailed etymologies – anyone can extend the list.

National traditions vary widely in this respect. For example, American English dictionaries pay great attention to synonyms. Noah Webster led the way in this respect. Even Webster’s Collegiate (a superb product in every respect) has always been keen on explaining the differences between words of more or less similar meaning, while in Britain discussion of synonyms has of old been the domain of specialized dictionaries, which tended to add etymologies. German, Scandinavian, and Russian “thick” dictionaries ignore synonyms, though there, too, specialized dictionaries exist, sometimes mere synonym finders, like the American Synonym Finder (Reference RodaleRodale 1961 and thenceforth). To make up for the deficiency in the treatment of synonyms, Germans were pioneers in compiling combinatory dictionaries, for it was they who developed a branch of linguistics that came to be known as Valenzgrammatik.

Dictionaries also tell us how to pronounce words, but, to convey phonetic information, lexicographers need transcription, and, though they have always made heroic efforts to facilitate the problem for the user, they failed, and failed rather signally. The problem is especially acute for English, with its erratic spelling of so many words. In French, ambiguous spelling is not uncommon, but most cases fall into several predictable groups (liaison before h, some final consonants, and so forth). The spelling of German and the Scandinavian languages is rarely too confusing. In Russian, the only severe problem is the placement of stress, and dictionaries have no trouble dealing with it.

Therefore, here we look at English. Consider the treatment of slough in several dictionaries (I dispense with references because it is only the result that matters). Slough is tricky because “mire,” “creek,” “despair,” and “a snake’s skin” have different pronunciations, and some differences between British and American norms exist. Here are the transcriptions I have found in representative dictionaries: ([slou, slö, sluf]; [slou, sluf] or [slau, slaf], the simplified version of the previous version; [slou, sloo, slŭf], [slo͞ o, slou, slŭf] and [slou, sloo, sləf]. The signs of the International Phonetic Alphabet (IPA) are avoided everywhere, but the results sometimes look even more exotic than the IPA’s [slou, slau, slʌf]. For slur dictionaries offer [slêr, slɜ(r), slur, slər]. Murray’s OED had its own set of signs. Wherever we look, we discover that, to consult even what purports to be a user-friendly transcription, one should learn and constantly remember a special alphabet.

It is also hard to decide how “broad” transcription should be. The concept of the phoneme is seldom called into question, but users need a clear guide to pronunciation, rather than a display of articulatory niceties or rigid adherence to some theory. English diphthongs are monophonemic. Yet, traditionally, two letters are used to transcribe them as in [sou] (so). Reference Kenyon and KnottJohn S. Kenyon and Thomas A. Knott (1944), distinguishing between cloth and clothe, wrote [klɔθ] and [kloð], but clout appears as [klaut]. I find their transcription [el] (ail) versus [ɔil] (oil) misleading and see little reason in having [peint] (paint), as opposed to [paint] (pint) and [pɔint] (point). After all, [pānt] (paint) and [pīnt] (pint), which we find in the Random House Dictionary (2001) and various editions of the American Heritage Dictionary, convey all the information one needs. The realizations of phonemes change over time, and soon after Jones, British dictionaries began to transcribe the diphthong of so as [ɛu]. In surveying this scene, one concludes that dictionaries sometimes try to chase the rainbow. Average users (especially foreigners) are puzzled by the spelling of bow. They will be grateful for the simplest tip: low bow is [lō bau], bowtie is [bōtī], and the like. The less philology or linguistics, the better. Professional phoneticians and phonologists should play on their own field because, unlike philologists, ordinary users do not seek (and may not benefit from) phonological precision.

Nowhere else does the dictionary depend more on the user’s level of preparation than in its treatment of etymology. Here, too, traditions vary. For example, Russian “thick” dictionaries sometimes indicate the source language of words still felt to be foreignisms but do not touch on the native stock. In Germany, Duden’s etymologies are short. The last German comprehensive dictionary that discussed word origins at length (Weigand-Hirt 1909–1911) appeared more than a century ago and is still valued mainly for this component. By contrast, in the English lexicographical tradition, etymology has always been prominent and remains so. Nathan Bailey’s dictionary (Reference Bailey1721) has appeared in countless reprints, and his modest etymologies commanded respect for a long time.

Above, mention was made of Webster’s contributions to the study of word origins. In 1864, the publishers invited Carl A. F. Mahn, a German scholar (and the author of many essays on word origins: see Reference MahnMahn 1855), to update, or rather rewrite, Webster’s explanations. This event was considered to be so important that the edition became known as Webster-Mahn. By contrast, it never occurred to anyone to refer to Webster-Skeat. (An aside: a similar event happened more than a hundred years later when the first AHD edition (1969) appeared. It was praised far and wide for its “Indo-European Roots” section, in which English words traceable to such roots were gathered in a special supplement (later published in book form [Reference WatkinsWatkins 1985]), and one could see at a glance that, for instance, Engl. flow, fluke, flake, flaw, plea, and a few others are certainly or possibly related. For contemporary lexicography it is a rare “victory” of philology over linguistics.)

Experience shows that presenting etymology to the uninitiated is an even more difficult task than providing them with understandable transcriptions. With the establishment of the comparative method, references to the material of entire language families became the norm (the present discussion is limited to Indo-European). It was taken for granted that the user would benefit from lists of multiple cognates, bear in mind the existence of sound correspondences, and appreciate the subtleties of Greek, Old English, Old High German, Old Icelandic, and Old French historical phonetics, to say nothing of asterisked (reconstructed) forms. (Against this background, it must perhaps have looked strange to the public that so many words were proclaimed to be of unknown origin.)

Before drawing conclusions about the state of etymology in non-etymological dictionaries, it may be useful to reproduce several samples. The first set deals with the origin of the English verb go.

• Wyld’s Universal: “O.E. gān, M. E. gō(n); O.H.G. gēn, gān; Swed. gå. Orig. a vb. in -mi; as appears fr. O.H.G. gām, ‘I go.’ Of the various suggested etymols. the most probable is perh. that wh. derives the word fr. an Aryan base *g̑ hei-, ‘to move, leap,’ & connects it w. Scrt. jíhītē, ‘he goes away, goes forward; leaps up,’ & further w. Lat. hēres, q.v. under its Engl. derivative heir; & with Lat. haedus, ‘he-goat.’ See further under goat.”

• AHD (Reference Morris1969): “Middle English gon, gōn(e). Old English gān, gegān. See ghē- in Appendix.*” The appendix, at ghē-, says: “To release, let go; (in the middle voice) to be released, go.” This is followed by 1. Suffixed form *ghē-no-, 2. Suffixed form *ghē-ro-, 3. Suffixed o-grade form *ghō-ro- (all with references to English words containing those roots), and four more suffixed forms, two of them containing schwa, and one beginning with perhaps. The entry ends with a reference to Pok. ghē- 418.

• Random House Dictionary of the English Language (RHD; 1987, repr. 1993): “[bef. 900; ME gon, OE gān; c. OHG gēn, G. gehen.”

• New Oxford American Dictionary (NOAD; Reference McKean2005): “Old English gān, of Germanic origin; related to Dutch gaan and German gehen.”

For comparison, this is what the same dictionaries say about the verb look:

• Wyld’s Universal: “O.E. lōcian, M.E. lōkie(n), lōken, cp. O.S. lōkōn, O.H.G. luogen, ‘spy out.’ See lynx.”

• AHD (Reference Morris1969): “[Middle English loken, to look, have the appearance, Old English lōcian, to look, from West Germanic lōkōjan (unattested)].”

• RHD (Reference Flexner1987): “[bef. 900, (v.) ME lōk(i)en, OE lōcian; c. MD lœken, akin to dial. G. lugen to look out.”

• NOAD: “Old English lōcian (verb); related to German dialect lugen.”

By way of postscript, all sources say about dodge: “Origin unknown.”

Even the briefest survey of the samples given above will show that the idea of including etymologies in dictionaries produced questionable results. The information we have seen is either highly professional but incomprehensible to anyone except a trained philologist, as in Wyld, or so meager and, at first sight, haphazard, as to be almost a waste. Out of context, the user will probably wonder why it is important to know the forms in Old Saxon and Old High German, how to “compare” look with Middle Dutch and a similar verb in a German dialect. And who was Grimm, the inventor of the term umlaut? Caesar, who coined ablative, is better known. The attempt to avoid asterisked forms, euphemistically called unattested, do not go far. AHD’s reference to the Indo-European roots in Julius Pokorny’s dictionary are confusing. We wonder again: who is that seemingly infallible expert? RHD and NOAD tell us that the verb go can be traced to Old and Middle English (the same holds for look). Why is this fact relevant?

Only one question interests the people who consult the sections on word origins in our dictionaries, namely (to cite the example given above), why the familiar sound complexes go, look, and dodge mean what they do. Such inquisitive people do not realize that outside indubitably sound-imitative and sound-symbolic (expressive, echoic) words, no one knows the answer and will hardly ever know it.

The situation looks less hopeless with borrowings. Here one can open the dictionary and learn that a certain word has its source in Latin or some other language, but even that information is of course not an etymology. Some time ago, lexicographers realized the dubious value of the etymological information they provide and went in two directions.

The most radical way was chosen by Funk and Wagnalls (1959). In the introduction (xvi), we read: “The etymologies, which were prepared under the direction of Professor Francis A. March, trace the meaning of any given word back in a direct line, avoiding mere guesses at derivation and temptation to make extended incursions into cognate languages.” This is a somewhat disingenuous way of saying that the dictionary was stripped of all etymological information. March did not leave a visible trace in comparative studies, but, judging by the reviews he published in 1881, he kept abreast of the development of etymology. In 1959, no trace of his input was left. Occasionally one runs into notes like ampula [L] (that is, Latin) and coin [F. < cuneus ‘wedge’]; F stands for ‘French.’ It seems that even such crumbs are the result of an oversight. Etymology has been a staple of the most influential English and German dictionaries for so long that one hesitates to celebrate its demise in them.

The other approach was to revise etymologies and make them understandable and interesting to lay readers. I think here the pioneer was AHD in its later editions. The editors added “word histories” to their popular and controversial panel discussions of usage. The novelty was a success, and other dictionaries imitated it, Merriam-Webster online among them. RHD hoped to bring out a new third edition of the great work, and I received an invitation to write a certain number of word histories for that never-completed dictionary; I still have about 200 of them in my archive.

Below, I reproduce three of those histories as they were supposed to appear in a third edition of RHD. They deal with an ancient culture word (write), a borrowing (madrigal), and a word whose mysterious origin was discovered with absolute certainty, a case rare in etymology (pedigree).

Obviously, my word histories would have supplemented, not supplanted, the etymologies in the dictionary. Therefore, in my text, references to Old Norse, Old English, Old Saxon, and other languages, along with abbreviations like ON and OE, looked natural. Besides, not unexpectedly, I chose only words I deemed interesting and those whose history I had researched before the invitation to write them for the never to be RHD3. It remains unclear how many such histories should be included, how long they should be, what to do with verbs like go and look, let alone prepositions and conjunctions. The aim “to avoid mere guesses at derivations and the temptation to make extended incursions into cognate languages” (FW) is admirable, but, unfortunately, there is little else in etymology. Funk & Wagnalls did throw out the bathwater with the result we have seen, and as the unpublished word histories for write, madrigal, and pedigree show, dictionaries can’t deal effectively with etymology without resorting to encyclopedism.

As regards linguistic terms, it is reasonable to include as few of them as possible in a general-purpose dictionary and to supply them with short, comprehensible definitions. A. S. Hornby solved some of the problems in this area well. Compare: “verb word or phrase indicating what sb. or sth. does, what state sb. or sth. is in, what is becoming of sth. or sb.”; “prefix word or syllable, e. g. pre- co-, placed in front of a word to add to or change its meaning”; “tense verb form that shows time: the present / past, etc.~”; “agglutinative (of languages) that combine simple words into compounds without change of form or loss of meaning.” Sometimes an obscure definition is saved by citing an example, as in “participle ‘hurrying’ and ‘hurried’ are the present and past ~s of ‘hurry.’” (For comparison, see what Wyld says about agglutination in his Universal Dictionary: “mode of combining together originally distinct elements in language, in which the fusion appears to be less intimate than in so-called inflexion, but which serves the same purpose as the latter.”)

23.5 Linguistics and Some Specialized Dictionaries

Mention should be made of the dictionaries focused on linguistics and philology per se. Etymological dictionaries are the most noticeable works of this genre. Even though before the discovery of sound correspondences, word historians, beginning with Plato and through the early nineteenth century, depended on lookalikes and on preconceived ideas of where words should, rather than do, come from, their efforts must not be forgotten, because they sometimes guessed well or knew something about the most recent coinages and borrowings we no longer know and because later researchers often reinvent mistaken derivations refuted by their predecessors. Perhaps in no other branch of linguistics/philology should one know everything about the chosen subject, and nowhere is it harder to amass a representative bibliography (but see my surveys in Reference LibermanLiberman 1998 and Reference Liberman1999). Fruitful ideas on word origins crop up in chance footnotes, newspaper articles, letters to editors, popular journals, and publications on any area (look for what has been said on madrigal in works on musicology and early poetic analyses and for explanations of pedigree in manuals of genealogy).

All this makes the existence of etymological dictionaries that offer exhaustive surveys of past scholarship so valuable. Here, suffice it to say that such dictionaries have been compiled (by first-rate specialists!) for Sanskrit, Classical Greek, Latin, partly for Hittite, and for several Slavic, Romance, and Germanic languages, but not for English. Though the OED provides etymologies, adding critical surveys of multiple hypotheses is not part of its mission. Elmar Seebold, the latest editor of Kluge, adds a few references here and there, but his choices are not always obvious. As a consequence, we need analytical dictionaries of etymology for specialists and serious amateurs, as well as word history notes for general dictionary users.

Etymology, like most of historical linguistics (“philology”), is a rare guest in today’s curricula, and jobs for specialists in this area are almost non-existent, even though books on word origins sell well and talks on the subject attract enthusiastic crowds. The great comprehensive etymological dictionaries of Hittite, Greek, Latin, Romance, Germanic, and Slavic mentioned above should be updated about every twenty years to keep up with scholarly advances, but few people are equal to the task (see Reference OguibénineOguibénine 2016: a rare example), and it is all but impossible to secure funding for such projects. Such is the state of the art.

Pronouncing dictionaries, too, are all about linguistics. Reference JonesDaniel Jones’ 1917 venture became a model for phoneticians in many countries in which, at first sight, they are not needed. In the Western world, nowhere else is spelling such an unreliable guide to pronunciation as in English (see what is said above about transcription). Yet pronouncing dictionaries have numerous features that justify the effort spent on their production. They list and rank variants (to the joy of many, they are descriptive, rather than prescriptive) and include countless place and proper names. And, of course, they are of inestimable use to non-native speakers.

Like other specialized works of reference focused on language, and to a greater extent than many, pronouncing dictionaries are hard to use without some knowledge of linguistics. It is not for nothing that in the last edition Jones (Reference Jones1957) lived to revise, he needed forty pages of explanations. Here are a few rubrics in his introduction: “Linking ‘r’ and intrusive ‘r,’” “Incomplete Plosive Consonants,” “Syllabic Consonants,” and “Strong and Weak Forms.” The volume concludes with a long glossary of phonetic terms.

Finally, of great importance are dictionaries of linguistic terms. Probably the earliest such dictionary was compiled by Jules Marouzeau and published in 1933. At the time, linguistics had not yet become a medley of conflicting schools. Later, each school produced its special doctrine and vocabulary. Today, even professionals are lost in the jungle of modern linguistic terminology. This is the justification of specialized dictionaries, even though some of them are doomed to become only monuments to short-lived theories.

For curiosity’s sake, it may be useful to look at the definition of philology in a few dictionaries of linguistic terms:

Reference MarouzeauMarouzeau (1951): “This word usually denotes the study of literature as such, but in a more narrow sense (except for historical disciplines: history and antiquities) the study of written monuments and the language forms they present, and in a more special sense the study of texts and their transmission, except for the research of the language, which is the sphere of linguistics. The English understand the term comparative philology as meaning comparative grammar or linguistics” (original in French).

Reference AkhmanovaAkhmanova (1966): “A general name of the disciplines devoted to the study of language, literature, and culture of a given people through literature and other works and monuments of culture and history” (original in Russian).

David Crystal’s Dictionary exists in several editions. Allowing for minor differences in print conventions, the first, third, fourth, fifth, and sixth editions (Reference CrystalCrystal 1980, Reference 674Crystal1991, Reference 674Crystal1997, Reference 674Crystal2003, Reference 674Crystal2008) defined philology thus:

The second edition (1984) and Crystal’s Encyclopedic Dictionary (Reference Crystal1992) offered a lengthier entry:

This text was abridged drastically, and the subsequent editions to date repeat the original (1980) definition.

We have learned a lot since the days of Charles Annandale, but the definition of philology remains the same, and the study of philology remains a noble, if unprofitable, pursuit. The subject “Linguistics and Philology in Dictionaries” has so many aspects and is so broad that the foregoing exposition does not pretend to be more than a first moderate approach to it, a mere invitation to develop this rarely investigated theme.

24.1 The Why of Lexicons

“The struggle of literature is in fact a struggle to escape from the confines of language; it stretches out from the utmost limits of what can be said; what stirs literature is the call and attraction of what is not in the dictionary.” So wrote Italo Reference CalvinoCalvino (1988), and he is right: literature is what happens after words find their right place on the page. But the quote fails to acknowledge that dictionaries are narratives (see, e.g., Reference AdamsAdams 2002a; Reference Considine and FontenelleConsidine 1998, Reference Considine2001, and Reference Considine2008). Without an obvious plotline, for sure, even disjointed, a list of words alphabetically organized, followed by their respective definition; yet each of those definitions is a story unto itself: not only about what that word says but about what it means and how it came to be that way. The understanding of love, for instance, is in constant change in dictionaries, and also of bad, patriotism, and home. In their aggregate, the definitions for these words and others represent a massive and methodical undertaking: to explain not only the vocabulary of a people at a specific time and place but their weltanschauung, too. Hence, any edition of the dictionary is a cultural snapshot – ephemeral, limited, and practical in its functions, just like the White Pages in telephone directories used to be: as much a glimpse of those who produced it as it is of those who use it.

The pitch I’m making isn’t a recent conception. While the voraciousness of the postmodern imagination, which looks at any artifact as a metaphor, has transformed dictionaries from reference tools to symbols of wisdom and even literary motifs, our predecessors were there already. Victorians read the New English Dictionary like a journal, maybe not literature in the sense “imaginative” literature, but not reserves or repositories of language either (see Reference Adams and GunzenhauserAdams 2010, 56–62). Likewise, paratexts of earlier dictionaries suggest intertextual reading admittedly on the part of the few. And modernists like T. S. Eliot, W. H. Auden, and others spoke of dictionary reading, and, in Auden’s case, of words themselves as poems (see Reference BrewerBrewer 2007, 190–191).

Plus, dictionaries are characters inside literature, which is again proof of Calvino’s owlish quotation: they show up in memoirs, novels, stories, poems, theater, and other genres, sometimes as passing references, at others as characters or as scaffolding that makes stories come alive. And they play another function: they are nation-building machines. It is impossible to imagine a nation without them simply because nations coalesce around a language and dictionaries serve as the mechanism to catalog that language. Indeed, dictionaries are gravitational forces around which literature, and culture in general, congeals; and the other way around, a national literature needs dictionaries to consolidate its existence as well as to test it (see Reference ConsidineConsidine 2008; Reference DollingerDollinger 2016; Reference GilliverGilliver 2016, 6–14). In other words, without a dictionary a nation is somewhat soulless. Freestanding dictionaries of the fully modern period are so implicated in nationalism that one can’t be without the other. That is the case of Robert Cawdrey, Henry Cockeram, Thomas Blount, or the glossaries at the ends of Renaissance books, etc. It might be said that as documents they aren’t dependent on an idea of “nation” in order to be written; still, they exude a national – and even nationalist – ethos. One can’t have dialect dictionaries until there’s a unifying identity against which to pose the dialects. Slang works somewhat the same way. Slang isn’t exclusively modern, as some claim, because it must position itself against a standard language, but slang dictionaries are counterpoised to the standard, which means they react to the national (see Reference ColemanColeman 2012, 57–58). Wherever a people come together as a nation, a dictionary is required to validate the intention. And it also becomes the symbol of amalgamation.

24.2 Idées Reçues

God, Maimonides argued, is the sole possessor of all knowledge. Every word past, present, and future is within the almighty. For humans, there is no such thing as an unabbreviated dictionary. We are trapped in our own narrow linguistic universe. Multilingualism is a way to escape that trap, but even the most expert of polyglots is limited. I’m familiar with the lexicographic enterprises of about a dozen traditions; Czech, Hungarian, Russian, Swahili, and Urdu, to list just a handful of important examples, are beyond my scope. The purview that follows is inevitably narrow.

Arabic philosophers and philologists in the Middle Ages, like Avicenna and Averroes, were interested in the study of alphabets. Al-Khalil ibn Ahmad al-Farahidi composed his Kitab al-‘Ayn, believed to be the first Arabic lexicon and arguably one of the first in any language, in the eighth century (see Reference BaalbakiBaalbaki 2019, 172). Yet dictionaries aren’t a theme in them. Nor are they present in the poetry of Hebrew poets of Spain during La Convivencia, such as Samuel ha-Nasi, Yehuda ha-Levi, and Shlomo ibn Gabirol, even though the first lexicographic reference in Hebrew, Saadia Gaon’s Agron, dates back to the tenth century (see Reference MamanMaman 2019, 186–187). (It also includes Arabic word translations.)

In France, Montaigne, in crafting his personal essays in the sixteenth century, patiently calibrates each of his words, often pondering their meaning. He will carve the way for Flaubert’s concept of le mot just, but again, as such, the dictionary is in the background. In Webster’s Third, Montaignesque is defined as “of, relating to, or having the characteristics of the essayist Montaigne, his literary style, or his thought.” Today there are dictionaries in French that survey Montaigne’s usage. His early education was in Latin. And he filled his Essais (1580) with foreign-language quotations. And he asserted that he had created a “dictionary all [his] own.”

Shakespeare might have invented around 1,700 English words (gossip, kissing, lament, undress, and zany). And the King James Version of the Bible, published in 1611, is a statement of his language. But the fact that the Bard makes no mention of dictionaries is in part due to the synchronisms of history. Ben Johnson’s words in “To the Memory of My Beloved the Author, Mr. William Shakespeare and What He Hath Left Us” argue in favor of a vast, sponge-like intellect with a limited education. There also weren’t any dictionaries of English available at the time until after Measure for Measure, probably written in 1603 and first performed in 1604. (It was published in the First Folio in 1623.) That leaves still a generous period of prolific output, since Shakespeare didn’t leave London until 1611. It’s possible, therefore, to think that Shakespeare was culturally uninterested in the useful if obnoxiously limited lexicons in circulation, or else that he used them but didn’t leave a record of it.

Miguel de Cervantes was also the creator of neologisms. In Part I of Don Quixote (1605), the priest and the barber visit Alonso Quijano’s personal library. The reader gets a glimpse of its content: mostly poems and plays. There is no dictionary, in part because, again, the first extensive one, Sebastián de Covarrubias’ Tesoro de la lengua castellana o española (1611), was still in the future (see Reference ConsidineConsidine 2019b, 302). In Part II of the novel, published in 1615, the knight and his squire enter a print shop during their visit to Barcelona. They find numerous works translated from foreign languages as well as an assortment of other books. But, alas, no reference to dictionaries here either.

In any case, it would not take long for European thinkers to conceive of the dictionary as a center of gravity. Although Daniel Defoe didn’t equip the protagonist of Robinson Crusoe (1719) with a dictionary, he surely was worried enough about the state of English in his time to propose the creation of an academy to protect it, the way the French had with their Dictionnaire de l’Académie fronçoise since 1634. So did Jonathan Swift, who invented a variety of “nonsense” languages in Gulliver’s Travels (1726), like those spoken in Lilliput and Blefuscu. He not only offered “A Modest Proposal” to end poverty in Ireland by selling poor children as food; in 1712, Swift also published “A Proposal for Correcting, Improving and Ascertaining the English Language,” a pamphlet in which he championed a policing institution made of experts whose job it would be to kindly offer advice on English usage. But while dictionaries of varying quality were common in Swift’s time, he also didn’t think of one as a kind of authority manual that could be packed in Gulliver’s luggage.

During the European Enlightenment, particularly among the French Encyclopédistes, the idea of corralling human knowledge, and then cataloging it in accessible fashion, became an obsession. Voltaire, Diderot, Rousseau, Jaucourt, and others were behind such universalist views. The universe, of course, was always being seen through strict European – and Europeanizing – eyes. It is noteworthy that the distinctions among glossaries, dictionaries, and encyclopedias, employed at times as synonyms, is a confusion ratified by this age: they are efforts at zooming on information; the decision how to organize that information is still ambiguous. The admirable dream of the Encyclopédistes was to look at words as implements to organize information, to bring the miasma of order into our chaotic surroundings.

Samuel Johnson, their contemporary and a magisterial man of letters, took the lead in ways still astonishing to us. No other lexicographer, regardless of the language, embarked on the effort to compile a dictionary of his own native language and largely succeeded. Johnson was also a literary scholar and Shakespeare specialist, a cultural critic, a travel writer, a novelist, and a poet. Interestingly, in his fiction, dictionaries don’t play an important part. Perhaps this is because, as he famously put it in A Dictionary of the English Language (1755), lexicographers, by his account, are “harmless drudges.” He was neither. Intriguingly, he defined novel as “a small tale, generally of love,” highlighting the degree to which only stories have depth and complexity outside of books.

To validate his definitions, Johnson, though not the first one, turned the effort of using quotations from literary authors, most prominently Shakespeare but also many others, into a validating tool. That strategy had been used in other European dictionaries, among them the six-volume Diccionario de Autoridades (1726–1739), composed during the reign of Felipe V, which made frequent reference to Spanish Golden Age authors like Cervantes, Lope de Vega, Góngora, and Quevedo. This strategy, among other things, consolidated the marriage between lexicography and literature: to be taken seriously, and for a nation to manifest its pride, lexicons showcased their respective writerly traditions.

Anatole France, author of Thaïs (1890), defined dictionary as “the universe in alphabetical order.” Flaubert led the way in looking at the structure of dictionaries as tools for literature. His Le Dictionnaire des idées reçues, compiled in the 1870s in journals in which he lampooned figures of his time, wasn’t published until 1911. Known in English as the Dictionary of Fixed Ideas, it is a kind of spoof encyclopedia through which he made fun of some of the platitudes of the Second French Empire. At some point, it appeared as if he intended his compilation to be an appendix of his posthumous novel Bouvard et Pécuchet (1881), which, somewhat like Diderot’s Jacques le fatalist et son maître (1796), uses narrative devices to tackle philosophical questions. This is one of the first instances in which the dictionary is presented as the structure to satirize an epoch.

It was in the late eighteenth and throughout the nineteenth century, the age of cosmopolitanism, that placed the dictionary at the center stage of culture. Writers no longer used it to consult; they now read it tirelessly and wrote about it as a subject of adoration. In America, Emily Dickinson’s niece Martha Dickinson Bianchi reminisced that her aunt read the dictionary (notably Noah Webster’s An American Dictionary of the English Language [1844]) “as a priest his breviary.” And essayist Ralph Waldo Emerson was a lover of dictionaries, which he approached as the inspiration for any literary work worthy of lasting across time, though in his work Emerson never turned them into raw material, at least not extensively. Emerson argued once that dictionaries are a good read. “There is no cant in it,” he stated, “no excess of explanation, and it is full of suggestion. The raw material of possible poems and histories.” While compiling the Oxford English Dictionary, James Murray, in 1895, used the comment as an illustrative quotation in the entry for the word dictionary.

Whitman is another superb case. His lexical range in Leaves of Grass (1855) was expansive. He wasn’t only fascinated with American slang but, judging from the magazine pieces he published and the notes and marginalia he kept, he intended to produce a dictionary. In one place, he wrote that “A Perfect English Dictionary has yet to be Written.” He believed the work of Johnson, Webster, Thomas Sheridan, John Walker, William Perry, John Ash, Nathaniel Bailey, John Kendrick, Benjamin H. Smart, and Joseph E. Worcester was a good start, “yet the dictionary, rising stately and complete, out of a full appreciation of the philosophy of language, and the unspeakable grandeur of the English dialect has still to be made.” He wanted his dictionary to account for philological issues as well as showcase the national character of the United States as perceived through its language (see Reference Blades, Pennington, Blades and PenningtonBlades and Pennington 2020, 13–14). But as such neither any real dictionary, nor the utopian version Whitman imagined, appear in his poetry.

In William Thackeray’s Vanity Fair (Reference 752Thackeray1848), set during the Napoleonic Wars, Johnson makes a cameo appearance. At one point, Miss Pinkerton writes her own name, and Miss Sedley’s, in the flyleaf of a Johnson’s revered Dictionary – “the interesting work which she invariably presented to her scholars, on their departure from the Mall.” Arguably one of the feistiest American writers, and a devotee of dictionaries, was the civil war poet, fiction writer, journalist, and American Civil War veteran Ambrose Bierce, known today for the extraordinary story “An Occurrence at Owl Creek Bridge.” Among his most famous volumes is The Devil’s Dictionary (1909), a lexicon in which Bierce sarcastically organizes knowledge alphabetically and as dictionary entries in order to criticize the mores of his time. Politicians and lawyers were Bierce’s principal targets. A capital was, in part, “The seat of misgovernment.” A lawyer was “one skilled in circumvention of the law.” And a liar, “A lawyer with a roving commission.”

The celebration of the dictionary as a linguistic fountainhead was, in Bierce, transformed into a cautionary tale. From the human cost of the Civil War to the upheavals in Russia and elsewhere, his was an age of skepticism toward politicians as benign leaders. Sarcasm spilled into other human endeavors, such as the objective compilation of knowledge. Bierce defines dictionary as “a malevolent literary device for cramping the growth of a language and making it hard and inelastic.” But he adds: “This dictionary […] is a most useful work.” And lexicographer, “A pestilent fellow who, under the pretense of recording some particular stage in the development of a language, does what he can to arrest its growth, stiffen its flexibility and mechanize its methods. For your lexicographer, having written his dictionary, comes to be considered ‘as one having authority,’ whereas his function is only to make a record, not to give a law.”

24.3 The Spinning Dictionary

Ezra Pound, an American expatriate who loved languages (aside from English, he knew Latin, Greek, French, Italian, German, Spanish, Chinese, Provençal, and Anglo-Saxon) and also dictionaries, published the volume ABC of Reading (1934), which uses the dictionary as a kind of structure. Despite its title, it isn’t about reading as such but about how to write poetry. He had tried the approach before. A year earlier, he published a collection of essays called ABC of Economics.

In 1940, Eliot, Pound’s friend, who was also born in the United States but became a British citizen, was asked if a great nation needed to have a great language, and if so, whether it was the business of the writer as artist to help to preserve and extend the resources of that language. Reference EliotEliot (Eliot and Hawkins 1940, 774/b) replied that “the dictionary is the most important, the most inexhaustible book to a writer. Incidentally, I find it the best reading in the world when I am recovering from influenza, or any other temporary illness, except that one needs a bookrest for it across the bed. You want a big dictionary, because definitions are not enough by themselves: you want the quotations showing how a word has been used ever since it was first used.” Eliot wasn’t thinking of the OED but of the Shorter Oxford.

As the various successive editions, as well as the supplements of the OED, appeared over time, the database of contemporary writers was updated. Authors like Auden, Virginia Woolf, and George Orwell were quoted. Woolf, in The Voyage Out (1915), attempted to write a novel about silence, one featuring “the things people don’t say.” Throughout her career, she sought a language that was beyond any of the ones available to humans, and along the way a dictionary capable of depicting happiness and longing far more accurately than with the words English has at its disposal. Meanwhile, Reference AudenAuden (1948, 4) argued that “though a work of literature can be read in a number of ways, this number is finite and can be arranged in a hierarchical order; some readings are obviously ‘truer’ than others, some doubtful, some obviously false, and some, like reading a novel backwards, absurd. That is why, for a desert island, one would choose a good dictionary rather than the greatest literary masterpiece imaginable, for, in relation to its readers, a dictionary is absolutely passive and may legitimately be read in an infinite number of ways.”

Conversely, this group of writers occasionally reflected on the dictionary as a source of control and as a receptacle of collective memory. Orwell (whose adjectivized last name, Orwellian, is defined as “characteristic or reminiscent of the world of Nineteen Eighty-four [1949], a dystopian account of a future state in which every aspect of life is controlled by Big Brother, by the British novelist George Orwell [1903–1950])” (Oxford Dictionary of Phrase and Fable), turns the dictionary into an image in his dystopian novel. In the back matter, he includes an extraordinary section on Newspeak, defined as “a language that sounds impressive but deliberately hides the truth and tries to change people’s traditional views about something” (Reference CareyCarey 2019). However, Nineteen Eighty-Four doesn’t include a lexicon per se.

By the second half of the twentieth century, the idea used by Pound and others to organize one’s thoughts and even one’s own life experiences in the manner of dictionaries took hold. Polish essayist and poet Czeslaw Miłosz also wrote a couple of autobiographical volumes, Miłosz’s ABC’s (1997) and A Further Alphabet (1998), in which he organized his life using the alphabetical sequence. So did Carlos Fuentes in This I Believe (2002), a dictionary of his ideological preferences. I’m not sure these narratives are actual tributes to dictionaries as such or to encyclopedias, which proves, in any case, that the line between these two formats remains muddled (see Hancher 2019; Liberman, Chapter 23, this volume).

For Jorge Luis Borges, dictionaries don’t occupy a prominent space in his oeuvre, but encyclopedias do. In the case of the encyclopedia of Tlön, of which it is said, in the story “Tlön Uqbar, Orbis Tertius” (first published in the magazine Sur [1940] and reprinted in Ficciones [1944]), that it is an idealized territory that only exists in the minds of a few initiated followers. Elsewhere, Borges writes about alternative languages, some of them created ex nihilo, for instance in “The Analytical Language of John Wilkins” (Other Inquisitions: 1937–1952), which argues that “[t]he dictionary is based on the hypothesis – obviously an unproven one – that languages are made up of equivalent synonyms.” In Wilkins’ arbitrary language, Borges states, he divided the universe “into forty categories or classes, which were then subdivisible into differences, subdivisible in turn into species. To each class he assigned a monosyllable of two letters; to each difference, a consonant; to each species, a vowel. For example, de means element; deb, the first of the elements, fire; deba, a portion of the element of fire, a flame,” and so on. Yet Borges doesn’t offer a dictionary of Wilkins’ enterprise, which would showcase its symmetrical structures.

In chapter three of Gabriel García Márquez’s One Hundred Years of Solitude (1967), an epidemic of insomnia sweeps over Macondo, the mythical town where the action is set. Soon the lack of sleep leads to another equally troublesome symptom: the loss of memory in the town population. José Arcadio Buendía, the founder of Macondo, devises a method made of tags – a kind of Montessori system – that allows people to remember what objects are for: “This is the cow. She must be milked every morning so that she will produce milk, and the milk must be boiled in order to be mixed with coffee to make coffee and milk.” Jose Arcadio Buendía’s idea is successful; it then leads him to his next invention: the memory machine. “The artifact was based on the possibility of reviewing every morning, from beginning to end, the totality of knowledge acquired during one’s life.” And here’s the pertinent aspect of it: “He conceived of it as a spinning dictionary that a person placed on the axis could operate by means of a lever, so that in very few hours there would pass before his eyes the notions most necessary for life.”

One of the most beautiful literary tributes to dictionaries is by Chilean poet Pablo Neruda. His “Ode to the Dictionary” is one of the 250 odes he wrote in the middle period of his career. In it he describes the ambivalent relationship he felt toward the dictionary when he was young and how, as years went by, it became an invaluable companion for him. I find it stunningly moving. Here is an English version, from Selected Translations: Poems 2000–2020 (Reference StavansStavans 2021, 98–100).

By the end of the twentieth century, postmodern games with the alphabet, and sometimes with the dictionary as well, were common in the sixties and beyond. Among the most famous is George Perec’s novel La Disparition (1969), known in English as A Void (1994). Perec writes the entire narrative without the letter “e,” quite a challenging feat given its invaluable status as one of the five most important letters of the alphabet. Needless to say, any translation of it – the English one, called A Void (1994), was done by Gilbert Adair – is equally difficult, as much a jigsaw puzzle as a full-fledged narrative. It isn’t surprising that Perec loved dictionaries. But as a presence the dictionary isn’t quite a character in his oeuvre.

A bit closer perhaps, though still resisting the presence of the dictionary as such, is another postmodern novel, Alphabetical Africa (1974), by the Jewish-American writer Walter Abish, which follows what has come to be known as an alliterative approach to literature. The architecture of the volume is interesting: each of the first twenty-five chapters is devoted to a different letter, containing only words starting with “a,” or “b,” and so on. Then the flow is reversed, letters are removed in the reverse order. The idea is intriguing, but is it good? Milorad Pavić’s Dictionary of the Khazars (1984), written in Serbian, is built as an encyclopedic dictionary about the Khazars, who supposedly converted to Judaism at the end of the eighth and beginning of the ninth century. The three sections in the novel are each written from another perspective from the viewpoint of the three Abrahamic religions: Judaism, Christianity, and Islam. Though it purports to be a historical game, the characters in the book are fictional.

Likewise, the use of the dictionary as a manual for specific emotions has become a standard trope. Han Shaogong’s A Dictionary of Maqiao (1996), set in a Chinese village in Hunan province, is compiled as a dictionary composed of 115 articles written by a student sent by a policy institute in the 1960s in the People’s Republic. The author was accused of imitating Pavić and went to court to defend himself. And Xiaolu Guo’s A Concise Chinese-English Dictionary for Lovers (2007) is a novel about a Chinese woman who has an affair with a British man. As she struggles to make sense of her emotions, she puts together a dictionary.

One writer who displays a fascination with dictionaries, encyclopedias, and with language as a semiotic code, is the Italian scholar and novelist Umberto Eco, author of the thriller set in the Middle Ages, The Name of the Rose (1981). Eco spent his life thinking about language as a system to organize reality. He published encyclopedias of various kinds, including one called On Beauty (2004) and another On Ugliness (2007). In Serendipities: Language and Lunacy (1998), he said that “the cultivated person’s first duty is to be always prepared to rewrite the encyclopedia” (Reference EcoEco 1998, 20), which, again, was a synonym of dictionary.

24.4 Dictionary Wars

The business of lexicography, its ins and outs, has spilled into mainstream culture. I have left until the end literary explorations with dictionaries at their heart. At the dawn of the new millennium, disquisitions on how words change, who is behind the adoption of any of them into the OED or Merriam-Webster’s dictionaries, and how dictionaries have come to occupy a privileged place in society are frequent topics of nonfiction books. Among the best-known is British journalist Simon Winchester’s The Surgeon of Crowthorne: A Tale of Murder, Madness and the Love of Words (1998), a history of how the OED was put together by professionals as well as lay contributors, one of whom was William Chester Minor, a surgeon in the US army who lived at the Broadmoor Criminal Lunatic Asylum, in Berkshire, England. My own Dictionary Days: A Defining Passion (2005) gathers personal reflections on the place of lexicons in my life. I have a collection of over two hundred dictionaries of all sorts of languages, ancient and modern, monolingual and polyglot, old and new editions. It is part of the Ilan Stavans Collection at the University of Pennsylvania. And Ammon Shea’s Reading the OED: One Man, One Year, 21,730 Pages (2008) is a marathon through the Oxford English Dictionary that demonstrates a lexicon also serves the function of a calendar. In embarking on his Quixotic quest, Shea discovers not only that he’s entered what Mario Vargas Llosa calls “a total narrative,” one encapsulating just about everything; he also realizes that the true protagonist of the OED is – surprise! – the reader.

Along the same lines, there are excellent volumes on how Johnson or Webster came to compose their dictionaries. Notable are Joshua Kendall’s The Forgotten Founding Father: Noah Webster’s Obsession and the Creation of an American Culture (2010), which places Webster in the context of his revolutionary contemporaries, including George Washington and Benjamin Franklin, and his dislike for Andrew Jackson. Jack Lynch’s You Could Look It Up: The Reference Shelf from Ancient Babylon to Wikipedia (Reference Lynch2016), full of juicy tales about figures like crackpot linguist John Horne Tooke, the magisterial OED editor James A. H. Murray, and British playwright George Bernard Shaw, who hopelessly sought to make English spelling less cumbersome. There’s also Simon Winchester’s The Meaning of Everything: The Story of the Oxford English Dictionary (2003), with Murray as arguably the OED’s most consequential protagonist. And Peter Martin’s The Dictionary Wars: The American Fight Over the English Language (2019), which argues that the rivalry between lexicographers like Webster and his more conservative nemesis Joseph Emerson Worcester, a traditionalist who was weary of language reform, came to symbolize in the nineteenth century the battle for the soul of America.

Again, there is more to say about German, Greek, Italian, Latin, Portuguese, Russian, and many other tongues. Each of these languages represents another national tradition. As I stated at the outset, there is no concept of nation in modern time without a language; likewise, every nation coalesces around a series of books, one or more of them a lexicon. Dictionaries might be seen as memory banks of lost civilizations. In what we know today as Latin America there were approximately 6,000 languages when Christopher Columbus arrived in 1492. A small fraction of them survives and an even smaller one has a dictionary, which could be seen as a tool of resistance. Another example is Yiddish, which, although almost completely obliterated by the Nazis during the Holocaust, refuses to die by continuing to produce lexicons. And Hebrew, a language resurrected from the dead, returned to life thanks, in large part, by the making of a modern dictionary by philologist Eliezer ben Yehuda, an effort that became the cornerstone of the Zionist movement in its early stages.

At any rate, the dictionary, pompous in its demeanor, mighty in its presentation, is, at its core, a rather volatile object. Authors do not always like them. In Elie Wiesel’s memoir Night (1956), as he looks for the right vocabulary to describe the atrocities he is witnessing at Auschwitz – he is writing in his mother tongue, Yiddish – “I would pause at every sentence, and start over and over again […] . All the dictionary had to offer seemed meager, pale, lifeless.” And in the novel The Wind-Up Bird Chronicle (1994), Haruki Murakami has the narrator state: “When I see a dictionary on my desk I feel like I’m looking at some strange dog leaving a twisty piece of poop on our lawn out back.” Dictionaries might be burnt, dismembered, and otherwise destroyed. Or else, they can simply be ignored, which is unquestionably the worst kind of censorship. Literary sequences in which books, including dictionaries, are obliterated, though they might not always be acknowledged, range from Elias Canetti’s Auto-da-Fe (1935) to Ray Bradbury’s Fahrenheit 451 (1953).

In short – and against Italo Calvino – the dictionary is where literature happens.

25.1 Introduction

“A good dictionary is essential to a manuscript editor.” So says The Chicago Manual of Style, often referred to as “the editor’s bible” (2017, 2.54). Likewise, as The Copyeditor’s Handbook advises, “A copyeditor must have a recent edition of a good dictionary – and must always keep this volume within easy reach” (2019, 69). Copyeditors would certainly rank among the most frequent users of dictionaries, consulting them on an almost daily basis. Editors turn to dictionaries for guidance on a variety of issues, from spelling and hyphenation to meaning and usage, even though the aim of dictionaries is simply to document words as they are used. Thus, the descriptions found in a dictionary become prescriptions in the hands of copyeditors, whether they seek to make a text conform to Standard English or simply to a prescribed style. The choices made by copyeditors may have significant effects on Standard English, as some linguists have noted: the editors of a recent book on the standardization of English wrote that “copy-editors control the gates to the world of print, and they play an important role in regulating the language” (Reference Pillière, Andrieu, Kerfelec, Lewis, Pillière, Andrieu, Kerfelec and LewisPillière et al. 2018, 18), while the founding editor of the popular academic journal English Today saw editors and proofreaders as “work[ing] at the coal-face of standardization” (Reference McArthurMcArthur 2001, 4). Perhaps it goes without saying that one of the primary tools that editors use in that regulatory role is the dictionary. Consequently, it is important to understand how editors use dictionaries and how editors and lexicographers reciprocally rely on and influence each other’s work.

The term editor may refer to several distinct roles, including acquisitions editors, who buy manuscripts and help shape their content; production editors, who manage and coordinate the publication process; and copyeditors, who find and fix problems in the text and sometimes prepare the manuscript for typesetting. Elsie Myers Stainton, former managing editor of the Cornell University Press, notes that “at some publishers one person may wear several of these hats” (Reference 750Stainton2002, 5). However, this chapter focuses on copyeditors, since their job makes use of dictionaries much more than the others. It also briefly discusses proofreaders, since their job also requires the use of dictionaries. Throughout the chapter, “editor” and “editing” refer specifically to copyeditors and copyediting unless otherwise noted. “Publisher” is also used broadly to refer not only to traditional publishers of books and periodicals but to companies and other organizations that employ copyeditors to help produce written works.

25.2 Why Copyeditors Edit

In this era of automated spellcheckers and grammar checkers, the need for editors is sometimes questioned. Pressures to cut costs have also led publishers to lay off in-house copyeditors and outsource the task to freelancers or to eliminate copyediting altogether, yet many publishers still find value in editing. Carol Fisher Saller, a former manuscript editor for The University of Chicago Press, writes of copyediting that publishers “obviously believe that it matters a lot […] because inaccuracies and inconsistencies undermine a writer’s authority, distract and confuse the reader, and reflect poorly on the company” (Reference Saller2016, 11). She adds, “Discriminating readers look for reasons to trust a writer and reasons not to. Sloppy expression and carelessness in the details are two reasons not to. […] We [edit] in order to help the writer forge a connection with the reader based on trust” (Reference Saller2016, 11).

Some editors view the needs of the reader as the primary motivation for editing. For example, the magazine editor and author Arthur Plotnik writes that “an editor’s only permanent alliance is with the audience” (Reference Plotnik1982, 25). Others take a broader view, as when Einsohn and Schwartz write that “copyeditors always serve the needs of three constituencies”: the author(s), the publisher, and the readers.

In addition to the obvious benefit of a livelihood, “Copyeditors are also rewarded, by the knowledge that something faulty was made acceptable, something good made better, and something very good made extraordinary. This knowledge translates into satisfaction” (Reference 750StaintonStainton 2002, 6–7).

Whomever it serves, the primary goal of copyediting is to facilitate communication from writers to their audiences; as Einsohn and Schwartz put it, the work of copyediting focuses on “the ‘4 Cs’ – clarity, coherency, consistency, and correctness – in service of the ‘Cardinal C’: communication” (Reference Einsohn and Schwartz2019, 3). That is, copyeditors strive to make sure a text is free from ambiguities and other problems that obscure meaning, that it holds together as a text, that it is uniform in matters of spelling and style, and that it is free from grammatical errors and nonstandard forms, as well as factual errors. They may fix awkward phrasing, make structures parallel, smooth transitions between topics, tighten up wordy passages, eliminate nonstandard or variant forms, and make the punctuation and capitalization conform to certain conventions, whether they are from an industry style manual or a house style guide. The aim of such changes is to ensure that readers understand what the writer is trying to communicate without unnecessary distractions.

25.3 The Copyediting Processes

Almost any type of text may pass through the hands of an editor, from independently published works to corporate communications to online content to traditionally published printed works. “Editors are everywhere,” Stainton writes: “They are at work wherever words are being written and published” (Reference 750Stainton2002, 3). After a manuscript has been acquired or a text has been written, it may undergo developmental or substantive edits by an acquisitions editor or someone in a similar role. These edits focus on the bigger picture, helping to shape the structure and content of the writing. Once the final manuscript is ready, copyediting begins, focusing more on the paragraph, sentence, and word level. The Chicago Manual of Style says that “manuscript editing, also called copyediting or line editing, requires attention to every word and mark of punctuation in a manuscript, a thorough knowledge of the style to be followed, and the ability to make quick, logical, and defensible decisions” (2017, 2.48). Editing may be done in-house or by a freelancer and is now usually done on screen using a word-processing software program such as Microsoft Word, which allows editors and authors to track changes, make comments, and convey queries.

Editing encompasses several distinct tasks, which may be done simultaneously or in different passes through a manuscript. Einsohn and Schwartz define six “principal tasks” (Reference Einsohn and Schwartz2019, 5) of the copyeditor:

• mechanical editing, which includes spelling, hyphenation, capitalization, punctuation, and other matters of style

• correlating parts, which includes cross-references, the numbering of notes and tables or figures, the contents of illustrations and captions, and the bibliography (Some manuscripts may require little or no work in this area.)

• language editing, which includes grammar, usage, and diction (Edits in this area are more subjective and require an understanding of which issues may confuse or distract readers.)

• content editing, which includes checking for internal inconsistencies and structural problems (The editor may also query facts that appear to be incorrect.)

• permissions, which involves noting any lengthy quotations or any tables, charts, or illustrations that have been previously published and require permissions or credits (As with correlating parts, some manuscripts may require no work in this area.)

• typecoding, which involves marking up different elements of the manuscript, such as headings, block quotations, and captions, for ease of typesetting and design. (Reference Einsohn and Schwartz2019, 6–12)

Editors may also do an initial cleanup of the electronic document, during which they fix any incorrect formatting, remove extraneous spaces or hard returns, and ensure that the proper characters are used for dashes, ellipses, and quotation marks and apostrophes.

Before editing begins, the editor considers factors such as the intended audience of the text, how the publication will be used, what level of editing is needed, what style is to be used, who the author is, and what deadlines exist. When time permits, the editor may skim or read through the entire manuscript before editing to get a sense of the scope of the work and to identify any issues that may need to be addressed before editing begins. An initial read-through should also give the editor a sense of how much editing is necessary, although instructions on how heavily to edit are likely also to come from the publisher.

Einsohn and Schwartz describe light, medium, and heavy levels of copyediting. At all three levels, the tasks of mechanical editing, correlating parts, permissions, and typecoding are the same; it is only language editing and content editing that vary. In a light copyedit, indisputable errors of grammar and usage are corrected, but wordy or convoluted passages, unfamiliar terms, and factual inconsistencies are merely queried. In a medium copyedit, the copyeditor also queries or corrects any infelicities in grammar and usage, suggests revisions to wordy or convoluted passages, asks for or provides definitions of unfamiliar terms, verifies content using standard references, and queries poor organization or logical problems. In a heavy edit, the editor corrects all grammar and usage problems, rewrites wordy or convoluted passages, verifies and corrects any factual errors, and queries or fixes any organizational or logical problems (Reference Einsohn and Schwartz2019, 14–19). By contrast with those three levels of editing, Chicago simply distinguishes between mechanical editing (addressing style, grammar, and usage) and substantive editing (resolving ambiguity and improving structure or organization) (2017, 2.49–2.50). Regardless of the level of editing, dictionaries will usually play a role in the process. In even the lightest, most minimalist edit, editors will consult dictionaries for guidance on issues like spelling, hyphenation, and usage.

Once editing begins, an editor typically makes two passes through a document, once to do the bulk of the editing work and once to catch any errors that escaped attention and to review or correct the initial edits (Chicago 2017, 2.53; Reference Einsohn and SchwartzEinsohn and Schwartz 2019, 23). After copyediting, the manuscript is usually returned to authors for review, at which point they can respond to queries, reject changes they disagree with, and make any other necessary adjustments. The manuscript is then typically returned to the copyeditor, who does one final pass to review author changes, ensure that all queries have been answered, edit any new text, and run the spellchecker one last time to catch any misspellings that are lingering or have been introduced later (Chicago 2017, 2.72). The manuscript then proceeds to design and then to proofreading and indexing.

Proofreaders look for some of the same issues that editors do, including spelling and hyphenation problems, the numbering of figures or tables, and missing or incorrect captions. Proofreaders also check end-of-line word breaks, typeface and font, page numbers and running heads, visual elements, and layout. They may also check for grammatical or logical problems, though changes in this area are generally minimal so as to avoid repagination, which may increase the cost of printing and introduce other problems (Chicago 2017, 2.110–2.118). During the final revisions, the changes from the proofread are checked, and the proofs are checked to make sure that no hyphenation or page break problems have been introduced.

25.4 Editors and Spellcheckers

The first time an editor actually uses a dictionary may be when they run the spellchecker, which some editors or style guides recommend as part of a first editing pass (see for example Reference Einsohn and SchwartzEinsohn and Schwartz 2019, 22). A spellchecker is essentially a type of automated dictionary, checking words in the document against words in the built-in lexicon. As Martin puts it, though, in the context of a word processor, “The user may not be aware that they are interacting with a dictionary […] at all” (2021, 219).

Spellcheckers are of course not foolproof; they can be tripped up by homophones, correctly spelled words that are not what the author intended (such as pubic for public), and different spellings determined by meaning (such as compliment and complement) (see for example Reference Einsohn and SchwartzEinsohn and Schwartz 2019, 141; MLA Handbook 2021, 1.14; Publication Manual of the American Psychological Association 2020, 4.30; Reference 750StaintonStainton 2002, 50). Spellcheckers may also accept variant spellings that do not conform to the relevant style guide or to the publisher’s preferences.

Some style manuals recommend using spellcheckers, though they also warn editors and proofreaders to watch out for errors that spellcheckers miss. Chicago, for example, says,

Einsohn and Schwartz note that “for long documents that contain many unusual words, spell-checking can be tedious,” and they warn against what are known as Cupertinos – errors introduced by autocorrect functions or “real words miscorrected into other words owing to a gap in the spell-checker’s dictionary, [or] misspelled words changed to the wrong word when the correct one is not listed as the spell-checker’s first suggestion” (Reference Einsohn and Schwartz2019, 172). (For more on Cupertino errors, see Reference ZimmerZimmer 2007.)

Spellcheckers may certainly be useful tools, but for these reasons editors do not fully rely on them and must still use other sources for spelling. Little or no research seems to have been done on the effect of spellcheckers on copyediting, though it seems reasonable to assume that spellcheckers prevent many misspellings from even reaching the copyediting stage. As Anne Curzan writes, Microsoft Word’s grammar checker – which includes a spellcheck function – is “arguably the most powerful prescriptive language force in the world at this point” (Reference Curzan2014, 64). She adds that “spelling tests, spellings bees, modern dictionaries, and the Microsoft spell checker all reinforce a belief that spelling does not and should not tolerate variation” (Reference Curzan2014, 73). Even though it is unclear precisely how or even whether spellcheckers have affected editors’ use of dictionaries, anecdotal evidence and professional experience indicate that editors still make regular use of dictionaries, not only for spelling but for a variety of other information.

25.5 When Editors Consult Dictionaries

Einsohn and Schwartz write that during the first editing pass, “you should look up anything you are unsure of,” adding, “With your dictionary, style manual, usage guide, thesaurus, and other reference books at your side, this is the time to read up on troublesome mechanical issues, brush up on tricky grammar and usage controversies, and verify your suspicions about factual inaccuracies or inconsistencies in the manuscript” (Reference Einsohn and Schwartz2019, 23). Editors may continue to use a dictionary during the second pass, as they check their work and look for any issues that may have slipped past them, or during the final review, as they check any changes from the author. And although Einsohn and Schwartz mention reference books, most editors nowadays use some sort of online or digital dictionary. Chicago notes that Merriam-Webster’s Collegiate is “regularly updated online” (2017, 7.1), and the latest editions of some dictionaries, such as Merriam-Webster Unabridged (2022) and The Oxford English Dictionary (2022), are available only online. Searching in an online dictionary is also faster and more convenient than using a print dictionary, and in some cases the online version may offer information lacking in the print version, such as extended usage notes or links to usage articles. (See, for example, the frequently asked questions and articles linked to below the entry for irregardless in the Merriam-Webster.com Dictionary [2022, s.v. irregardless].)

Wherever they may use them, editors consult dictionaries for a variety of reasons. The Associated Press Stylebook recommends that editors use a dictionary “for spelling, style and usage questions not covered in this stylebook” (2020, s.v. dictionaries), while the Publication Manual of the American Psychological Association refers editors to dictionaries in cases of spelling (2020, 6.11), hyphenation (6.12), some capitalization issues (6.14), treatment of foreign words (6.22), and abbreviations (6.25). Chicago is even more thorough, recommending that editors consult a dictionary for general spelling questions (2017, 7.1), plural forms (5.13, 7.5), comparative and superlative forms (5.87, 5.167), regular and irregular verb forms (5.100), the use of the indefinite articles a or an before words starting with h (7.32), end-of-line word division (7.36), identifying whether foreign words and phrases should be italicized (7.53), hyphenation (7.81, 7.89), compounds (7.82), abbreviations (10.1), and more. Einsohn and Schwartz give a similarly thorough list of information that an editor might look for in a dictionary:

Most editors will likely look for scientific names and biographical or geographical information in other sources first, and some editors may turn to a usage dictionary (see Chapman, Chapter 16, this volume) or a similar guide on questions of usage. Nevertheless, editors still use dictionaries for a wide range of information.

In addition to outright misspellings (such as definately for definitely) or nonstandard forms (alright for all right), editors will also watch for spelling variants. Although a dictionary may list multiple standard variant forms, style manuals often provide guidance on how to select among those forms. If the style manual does not provide such guidance, or if the publisher allows for the author’s preference or does not have a house style, the editor at least ensures that the author has been consistent in the choice of variants (Reference Einsohn and SchwartzEinsohn and Schwartz 2019, 7). Hyphenation and compounding are also a frequent source of dictionary lookups; Chicago says, “Far and away the most common spelling questions for writers and editors concern compound terms – whether to spell as two words, hyphenate, or close up as a single word.” Chicago adds that “the first place to look for answers is the dictionary” (2017, 7.81). Einsohn and Schwartz similarly write,

Benjamin Dreyer, the copy chief of Random House, also recommends consulting a dictionary for matters of hyphenation, though in somewhat tongue-in-cheek fashion: after noting the inconsistency between dictionary entries such as “light-headed” and “lighthearted,” he says, “If you’re invested in getting your hyphens correctly sorted out in compound adjectives, verbs, and nouns, and you like being told what to do, just pick up your dictionary and look ’em up. Those listings are correct” (Reference Einsohn and Schwartz2019, 56).

Proofreaders also use dictionaries for some of the same issues, ensuring that spelling, hyphenation, and compounding are consistent throughout the text and that they conform to the dictionary of choice. Likewise, they also make sure that words broken across lines follow the breaks indicated in the dictionary. They may also consult a dictionary for grammatical or usage questions, though these are usually infrequent at this stage.

But even when an editor or proofreader turns to a style manual for guidance on spelling, hyphenation, or compounding, the style manual itself often appeals to the dictionary. For example, Chicago spends several sections on hyphenation and compounds (from 7.81 to 7.89), including a table that spans more than eleven pages enumerating different kinds of compounds and the guidelines for their treatment. In these sections, Chicago’s dictionary of choice, Merriam-Webster’s Collegiate, is mentioned twenty times, eighteen of which occur in the hyphenation table, with several other generic references to dictionaries. For editors, dictionaries are not simply descriptive reference works but authorities and arbiters of correctness, and this view presumably stems partly from the widespread view of dictionaries as guardians of the language and partly from the way style guides themselves treat dictionaries.

25.6 Style Guides on Dictionaries

Style guides typically recommend or prescribe specific dictionaries, and these recommendations serve as the basis for the style guide’s preferred spelling and compounding style, with the style guides providing additional information, clarifications, or exceptions. But style guides also give explicit instructions on how to use those dictionaries, especially when it comes to variation. Chicago, for example, says,

The sixth edition of the Publication Manual of the American Psychological Association (APA) gives similar advice:

However, the more recent seventh edition has softened its stance somewhat and now says, “If the dictionary offers a choice of spellings, select one and use it consistently throughout your paper” (2020, 6.11). Though it recommends a different dictionary, The Associated Press Stylebook gives advice similar to that found in Chicago and previous editions of the APA style manual:

The MLA Handbook, by contrast, does not prescribe a particular dictionary, but it still recommends that users pick one dictionary: “To ensure consistency, use a single dictionary, such as Merriam-Webster (merriam-webster.com); if an entry has variant spellings, generally adopt the spelling listed first” (2021, 2.2). But a post on the MLA Style Center page says that “MLA publications generally follow the spelling preferences listed in Merriam Webster’s Collegiate Dictionary, available online, and Merriam Webster’s Unabridged, available by subscription” (Reference GibsonGibson 2016).

When a style guide does choose a particular dictionary, the choice can sometimes send a message about its stance on language. For example, some people were hesitant to adopt the controversial Webster’s Third New International Dictionary (1961) because of its alleged permissiveness, causing some publishers to reevaluate their choice of dictionary. Calling Webster’s Third “a flexicon that tossed linguistic standards to the winds and allowed slang to lie down with formal English,” William Safire wrote that The New York Times had decided to stick with the second edition – Webster’s New International Dictionary (1934) – “until the dictionary-makers came to their senses” (1975). Safire noted that as years passed, copies of the second edition became harder to find and, of course, increasingly out of date. The editors of The New York Times decided to adopt a new dictionary more aligned to the values of The Times than Webster’s Third, and they chose Webster’s New World Dictionary, published by Collins-World (1970). Webster’s Third was, however, retained as an auxiliary reference for words not found in the New World.

To exemplify the dictionaries’ different approaches, Safire discusses their treatment of the word kudos: “In Merriam‐Webster, ‘kudo’ is accepted as a singular backformation from ‘kudos,’ perceived as plural; but in Mr. Guralnik’s New World, ‘kudos’ is defined as ‘credit or praise… sometimes wrongly taken as a plural of an assumed ‘kudo’’” (1975). For at least some editors turning to a dictionary for guidance on points of changing or divided usage, Webster’s New World provided more authoritative rulings than Webster’s Third. Other news organizations soon followed the lead of The New York Times. As Webster’s New World explains, “The next year, both The Associated Press and United Press International adopted the dictionary and based their style manuals on it as well” (2014, vii). (See Finegan, Chapter 19, this volume). Nearly fifty years later, the Associated Press still prefers Webster’s New World, even though the furore over Webster’s Third has died down and indeed may be unfamiliar to younger editors.

In some cases, though, a dictionary is just a dictionary. Even though Chicago recommends Merriam-Webster’s Collegiate, it adds, “For further definitions or alternative spellings, refer to another standard dictionary such as the American Heritage Dictionary of the English Language. At least for spelling, one source should be used consistently throughout a single work” (2017, 7.1). That is, some publishers or clients may prefer a different dictionary, or editors may consult several dictionaries in the course of their work, but they should always use the same one to decide between spelling variants. Saller acknowledges that the choice of dictionary may vary according to preference or project: “Style rules (which pertain to punctuation, capitalization, hyphenation, preferred spellings, and conventions for citing sources, among other things) are often by nature arbitrary and changeable. If style rules were universal and immutable, there would be no need for different style guides and dictionaries” (Reference Saller2016, 33). Nevertheless, the implicit message seems to be that the choice between variant forms, and even the choice of which dictionary to consult regarding those variants, is ultimately less important than the act of choosing. Choosing a dictionary and a method for deciding between variants allows an editor to impose consistency on the text without spending too much time making those decisions.

Curiously, though, the need for consistency is rarely justified. In her book Verbal Hygiene, the linguist Deborah Cameron notes that the index to Stainton’s book The Fine Art of Copyediting has eleven entries on consistency, but none of them provide a rationale for its importance. When Cameron pressed her copyeditor informants for a reason, they advanced two arguments: first, that readers demand consistency and are irritated by inconsistency, even when one variant is intrinsically no better than another; and second, that inconsistencies distract readers from what is being communicated, effectively making the text less clear, even if that distraction is subliminal (Reference Cameron2012, 37). Cameron is skeptical of both arguments, writing, “Uniformity and transparency of style, especially the former, are essentially products of the craft tradition itself. To the extent that readers demand them, it is because they have been trained to demand them by consuming the products of craft professionals” (Reference Cameron2012, 38).

She also notes the irony of copyeditors looking to descriptive reference works like dictionaries for consistency; one dictionary may list one form first, while another dictionary may list another form first, and both will presumably list both variants. If dictionaries vary among themselves, and if they also describe the variation found in the language, then why treat them as arbiters of correctness or as tools for ensuring consistency? Cameron says that there are two reasons for treating dictionaries as authorities: “First, that despite their rhetoric they are really prescriptive, not descriptive; and second, that the differences in what they prescribe have nothing to do with the so-called ‘facts of English usage,’ but are … motivated by considerations of marketing and ‘brand image’” (Reference Cameron2012, 49). She concludes, “Variation is kept within reasonable bounds only because commercial rivals share a common interest in maintaining the belief that there is some uniform standard for correctness in language” (Reference Cameron2012, 49). Although it is clear that editors treat dictionaries as arbiters of correctness or consistency, the idea that dictionaries are linguistic authorities with an interest in maintaining standards of correctness stands in marked contrast with how lexicographers generally view their work.

25.7 Dictionaries as Editorial Authorities

Since the publication of Webster’s Third New International Dictionary, “descriptive and not prescriptive” (Reference Gove, Sledd and EbbettGove 1962, 74) has become a sort of unofficial slogan of English-language lexicographers, though this philosophy seems to have had little impact on how users view dictionaries. As Edward Finegan says, “We tend to rely heavily on dictionaries for definitive pronouncements about what is correct” (Reference Finegan1980, 3). James Milroy and Leslie Milroy even argue that dictionaries further the standardization process by accelerating prescriptivism: “Prescription becomes more intense after the language undergoes codification … because speakers then have access to dictionaries and grammar-books, which they regard as authorities” (Reference Milroy and Milroy2012, 22). And Deborah Cameron says that dictionaries “particularly invite questions about the nature of their authority, because that authority is so visible and so fetishized” (Reference Milroy and Milroy2012, 49). But lexicographers’ intentions may ultimately be irrelevant; as John Joseph argues, “The meaning of prescriptive and descriptive are in the use of your linguistic work – how it is interpreted and applied – rather than in your intention as an analyst, which no one else can know, only infer, in ways that will vary according to their own experience” (Reference Joseph2020, 22). That is, even as lexicographers strive to document words as they are used, editors infer their own meanings from dictionaries and consult them to determine how words should be used.

The APA style guide, for instance, says, “APA uses Merriam-Webster’s Dictionary and the APA Dictionary of Psychology as its authorities on spelling” (2020, 4.30). And editors often speak of the first variant listed in a dictionary as the preferred form, as if lexicographers are recommending it over other forms. Einsohn and Schwartz, for example, write, “Your dictionary will show one spelling as the primary variant, or preferred spelling, another as a secondary variant or as a British variant. […] A separate entry for a variant spelling, if provided, will refer you to the preferred spelling” (Reference Einsohn and Schwartz2019, 145; emphasis in original). Stainton says that “American dictionaries list theater as preferred and theatre as acceptable,” as if the ordering of variants denotes different degrees of correctness or at least favoredness (Reference 750Stainton2002, 64). And although many publishers and editors take a flexible approach, allowing for authors’ preferences or different choices of dictionaries, they still treat the dictionary as an authority to be used in defense of an editor’s changes. Saller, for instance, tells editors, “Always know why, according to an authority, you are making a change. […] Name the style guide and dictionary you used as arbiters in questions of consistency” (Reference Saller2016, 40). She also recommends that editors “consult an up-to-date dictionary,” in part because “the usage notes and discussion, sometimes extensive, settle many popular debates” (Reference Saller2016, 60). And Stainton reminds editors, “You have your press behind you and all the dictionaries and manuals of style” (Reference 750Stainton2002, 27). For editors, dictionaries do more than simply document the language – they tell editors which variants are most acceptable, they settle arguments, and they lend authority to editors’ decisions. The descriptions in a dictionary help motivate and justify an editor’s prescriptions.

Many linguists and lexicographers have noted the difficulty in disentangling descriptivism and prescriptivism. Don Chapman and Jacob Rawlins, for instance, call the prescriptivism-versus-descriptivism dichotomy “an untenable binary” (Reference Chapman2020, 5), while John Joseph says, “it is unclear whether pure descriptivism is possible,” arguing that “descriptions can in fact contain value judgments, and perhaps cannot escape doing so when it is human behavioural norms that are being described” (Reference Chapman2020, 19, 28). According to Anne Curzan, the very fact that dictionaries attempt to describe words as they are used allows users to appeal to dictionaries’ authority in determining what is and isn’t a “real” word. If a word is not in wide enough use to have been documented in a dictionary, then its absence may well be an indication that the word is outside the standard language or is “peripheral” in some way; she says that for most speakers, “real words are in dictionaries, without usage labels. Words not in dictionaries have a questionable status as words at best, and usage labels such as ‘nonstandard’ can shake speakers’ sense of a word’s full legitimacy” (Reference Curzan2014, 93–94). Entrance into a dictionary is also taken as evidence that a word is “accepted” – at least by that dictionary – as when Safire complained that Webster’s Third had “accepted” kudo as a singular form of kudos, when in reality it had merely recorded its use as a singular form.

Lexicographers are well aware of the tension between how dictionaries are made and how they are used. At the 2021 conference of the American Copy Editors Society, Peter Sokolowski, editor-at-large for Merriam-Webster, acknowledged this tension when he said, “The presumption of any dictionary is a prescriptive one. It’s absolutely telling you what the commonly accepted pronunciations and spellings are of a given word… . Its function is prescriptive, but its method is often descriptive” (Reference Sokolowski and BrewsterSokolowski and Brewster 2021). Descriptive works are taken as prescriptive ones partly because language descriptions, especially descriptions of standard or prestige varieties, help to create and enforce language norms. As James Milroy says, “All language descriptions, no matter how objective they are, must be normative. […] To be accurate they have to coincide as closely as possible with the consensus norms” (Reference Milroy1992, 8–9; emphasis in original). He also notes that these objective descriptions are interpreted prescriptively by language users, at least when these works describe the standard variety; he says that even a descriptive work like An English Pronouncing Dictionary (1955) is seen by its users as a guide to correctness because the variety it describes is “one that is desirable to acquire” (1992, 9). Responding to this idea, Cameron says, “Because science itself has authority in modern society, while at the same time the discourse of value remains a highly salient one for everyday talk about language, the absolute distinction between observing norms and enforcing norms cannot be maintained in practice” (Reference Cameron2012, 7–8). And, of course, even the most descriptive dictionaries are still copyedited and proofread before publication, showing just how inescapable norm enforcement is.

Even though editors and others may view the ordering of variants as an indication of which forms are more preferred or more accepted, lexicographers still take a more dispassionate view of their work, as evidenced in dictionaries’ front matter. As Giuliana Russo says, “Grammars and dictionaries usually reveal their theoretical background, and consequently their ideology and values, in their prefaces” (Reference Russo2020, 256). Einsohn and Schwartz recommend that editors read their dictionary’s front matter (Reference Einsohn and Schwartz2019, 71), but the front matter is sometimes difficult to find online. In the online American Heritage Dictionary, the front matter can easily be found by clicking on a link labeled “Guide to the Dictionary” in a sidebar titled “How to Use the Dictionary” (2022), but the front matter of the online Merriam-Webster.com Dictionary is harder to find; users have to scroll to the bottom of the page and click “Help” in the footer to find materials like the pronunciation guide and the explanatory notes to the dictionary (2022). And even if editors read the front matter, the ideologies and values found there do not square with how editors interpret – or are instructed to interpret – dictionary entries. Merriam-Webster’s Collegiate, for example, says:

The explanations of usage labels likewise take a rather neutral stance:

Webster’s New World, chosen by newspaper editors in protest over Webster’s Third, is similarly neutral in its descriptions of variants:

1. 1. as joint main entries: this implies that both spellings occur equally, or almost equally, often and that neither one is “more correct” or to be preferred… .

2. 2. at the end of an entry block, part of speech, or individual meaning, in small boldface: this treatment is used when the alternative spelling occurs less often than the main entry spelling or when it has a particular quality that needs mentioning, such as being British, dialectal, poetic, or rare. (2014, xiii)

Even The American Heritage Dictionary, which “began as a carefully imagined and articulated prescriptive response to what were seen by many Americans as the descriptive excesses of Webster’s Third” (Reference 697AdamsAdams 2015, 42), is scarcely more prescriptive than Merriam-Webster’s Collegiate or Webster’s New World in recent editions:

Regardless of lexicographers’ intentions, and despite their explicit guidance in dictionaries’ front matter, copyeditors routinely read dictionary entries prescriptively.

Copyediting thus provides an excellent example of how a dictionary’s observation of norms is easily turned into enforcement of norms. Major style guides typically prescribe a particular dictionary, and even when they allow authors or editors a choice of dictionary, they at least recommend that editors use a single dictionary for consistency’s sake. Most style guides also recommend that writers and editors use the first variant listed or the form listed in a main entry rather than a cross-reference. Whether editors infer prescriptions from a dictionary where none are intended or whether style guides explicitly instruct editors to use them prescriptively, dictionaries nevertheless become prescriptive tools in the hands of editors.

25.8 How Dictionaries Influence Editors and Editors Influence Dictionaries

When a word has variant spellings or variant forms, dictionaries typically list the more common one first, and, as stated above, the advice in most style manuals ensures that many editors will choose that form, reinforcing its dominance in published writing. As Joseph notes, grammarians and lexicographers rely heavily on literary citations, and “the examples at their disposal are with few exceptions the outcome of previous sifting by one or more editors” (Reference Joseph1987, 114), thus creating something of a feedback loop. Editors and lexicographers are at least somewhat dependent on each other, as each group examines the other’s work in order to determine either what norms to describe or what norms to prescribe. This feedback loop has not been explored in much depth, though Einsohn and Schwartz acknowledge its existence after a discussion of the dictionary’s “preferred” forms:

Cameron also noted the potential for feedback loops between dictionaries and their users:

Some dictionaries may also be more prone to such feedback loops than others. For example, the first edition of The American Heritage Dictionary relied heavily on the Standard Corpus of Present-Day Edited American English (the “Brown Corpus”). As Michael Adams writes, this “means that there is considerable bias in the construction of the dictionary towards copy-edited prose, where form and meaning have already been mediated in the direction of correctness” (Reference 697Adams2015, 31). Any other dictionary that relies on edited writing, whether from printed materials or from any of the other large corpora now available, has the same potential problem – it describes usage that has been filtered by copyeditors who consult dictionaries to determine what usage to allow in print.

Though little research has been done in this area, there is some evidence that editors’ preference for the primary variant may be contributing to some shifts in usage in standard written American English. For example, toward and towards appear in roughly equal frequency in unedited writing, but American editors change most instances of towards to toward, creating or at least contributing to the appearance that toward predominates in American English (Reference OwenOwen 2020, 300–301). This prevalence in published writing then sometimes makes its way into dictionaries’ usage advice, as in this note under the entry for toward in The American Heritage Dictionary: “Some critics have tried to discern a semantic distinction between toward and towards, but the difference is entirely dialectal. Toward is more common in American English; towards is the predominant form in British English” (2011, s.v. toward). Though editors are already typically instructed to use the first form listed, the idea that there is both a frequency difference and a dialectal difference between toward and towards may provide extra justification in editors’ minds for favoring toward, though further research would be needed to assess editors’ attitudes and practices in this area.

Occasionally editors lament the inconsistency between forms found in the dictionary (as with Dreyer’s complaint about light-headed and lighthearted). They may wish to adopt a different spelling or styling but feel beholden to the dictionary’s first or only listed form. For example, Mark Allen, a freelance editor and writer, wrote in 2011:

(As of this writing, though, the first form listed in Merriam-Webster online is goodbye, so Allen’s preferred unhyphenated form now has Merriam’s blessing.) But since dictionaries simply follow the usage found in published writing, editors’ decision to use the dictionary’s first form may prevent or at least slow the adoption of the forms they prefer. Sokolowski noted this problem at an American Copy Editors Society conference: “The authority, my friends, editors, is you, ultimately. The usage that we’re recording is often determined by the gatekeepers of publishing. The fact is we’re looking to you just as you’re looking to us, so there is a feedback loop. Sometimes that leads to slow changes” (Reference Sokolowski and BrewsterSokolowski and Brewster 2021).

Bryan Garner similarly noted the effect of lexicographers’ work on published writing: “If the work is done well, it has a normative influence within literary circles. Although dictionaries are supposed to reflect usage and describe usage, they inevitably influence usage. The effect becomes a cause” (Reference Stamper, Garner and GreeneStamper, Garner, and Greene 2021, 197). He illustrated his point with an example of shifting usage driven in part by a change in The AP Stylebook, itself a kind of dictionary (though an overtly prescriptive one):

In response, Stamper said, “This is a reality that as lexicographers we skirt around […]. We don’t reckon with the feedback loop that also editors are looking at dictionaries for guidance” (Reference Stamper, Garner and GreeneStamper et al. 2021, 199).

This feedback loop is, of course, to a large extent inevitable. Normativity is “an inalienable part of using language” (Reference CameronCameron 2012, 2) because users of the language naturally tailor their use to the norms of their community or interlocutors. This use then becomes part of the model on which other speakers base their use. But this normativity is strengthened as lexicographers work to document language norms and as this documentation is then taken as a prescription for certain language norms over others. Language users who are thus influenced by the dictionary then create more evidence of these norms through their own use, which dictionaries then document, thus reinforcing the norm. And, as said before, copyeditors are especially likely to view dictionaries as authorities; further, their role as gatekeepers of publishing gives them an outsized influence on what appears in print – and thus what is codified in dictionaries and therefore regarded as standard or correct.

25.9 Conclusion

Copyeditors and lexicographers rely on and influence one another’s work, with significant implications for both parties. It has long been recognized that dictionaries play an important role in standardization; Peter Trudgill says that language codification is “usually enshrined in dictionaries and grammar books” (Reference Trudgill1992, 17), and John Joseph calls dictionaries “the tool of standardization having the widest distribution throughout every level of the society” (Reference Joseph1987, 161). And through copyeditors, this standardizing role may be larger than previously acknowledged. Joseph says the influence of editors “has scarcely been taken into consideration,” adding,

Because editors exert such a powerful gatekeeping role, lexicographers may want to reconsider which sources they use and how they use them in creating and updating dictionaries. The usage found in published writing and in corpora of edited materials is not necessarily reflective of the usage of English speakers as a whole; in some cases it may be skewed in the direction of editors’ preferences, and some of the usage of other speakers could be underrepresented or even missing. Fortunately, new corpora have made it easier to find a wide variety of text, some formally published and some unedited and informally published, and it is possible to collect data from many other sources as well. Determining how to use these other sources is not without its challenges, of course. As Emily Brewster, a lexicographer with Merriam-Webster, said at the Annual Copy Editors Society conference in 2021, “Now anyone can have a platform… . The access that lexicographers have to the language of so many more speakers is just confounding” (Reference Sokolowski and BrewsterSokolowski and Brewster 2021).

Copyeditors and especially the authors of the style guides they follow may also want to be more cognizant of the fact that dictionary entries do not prescribe or express preferences in the way that many users assume. Some style guides have already taken a less rigid approach, allowing for the choice of dictionaries or allowing variant forms as long as consistency is maintained. Editors and style guide authors should also recognize that dictionaries are simply documenting the usage that editors allow into print; thus, if they want to see certain forms adopted or listed first in their dictionary of choice, they should use their power as gatekeepers to allow those forms in print rather than waiting for the dictionary to take the lead. There is, of course, still much research to be done on copyeditors’ use of and attitudes toward dictionaries, as well as the influence of copyediting on lexicography and of lexicography on copyediting.

26.1 Introduction

Law and lexicography intersect in various ways, involving intricacies in legal phraseology and semantics, the creation of dictionaries both legal and general, the evolving philosophies of what dictionaries should do, how lawyers and judges use dictionaries in their work, and even the settled doctrine that the drafters of legal instruments can be their own lexicographers. The two disciplines are perhaps as closely intertwined as any disparate disciplines can be – all the more so given the dramatic rise of textualism, especially in the United States, since the mid-1980s. Although a famous judge once cautioned that it is a mistake for advocates and judges to make a fortress out of the dictionary,Footnote ¹ that view has receded in recent years. If not a fortress, the dictionary is certainly a mainstay in modern judicial decision-making.

Here we’ll examine (1) the history of legal lexicography in Anglo-American law, (2) the methods of modern legal lexicography, (3) how courts use dictionaries, and (4) how legal drafters engage in nonce-lexicography when preparing legal instruments.

26.2 A Short History of Legal Lexicography

Although early legal lexicography is often slighted in historical studies of dictionary-making, the story of early legal lexicography in the English language, during the sixteenth and seventeenth centuries, is largely the story of English lexicography – which is to say that lawyers were greatly responsible for advancing the art. Lexicography, of course, is the writing of dictionaries – not “compiling,” but “writing.” Those two undertakings are very different. Proper dictionaries are composed, not assembled. Writing good definitions is a constant challenge for lexicographers of any kind.

Even if legal lexicography trailed a bit chronologically, it led qualitatively. The first monolingual general-English dictionary, Robert Cawdrey’s Table Alphabeticall (1604), started the tradition of Jacobean hard-word lexicons. It predates the first English-only legal dictionary, John Cowell’s The Interpreter (1607), by only three years. But Cawdrey’s Table runs to a mere 109 pages of defined terms, listing just over 2,500 words, each glossed with a synonymous word or phrase (almost always on the same line) – barely a dictionary at all. The average length of an entry is 9.2 words. By comparison, Cowell’s 557-page Interpreter contains more than 2,200 fully fleshed-out entries that make it immediately recognizable to modern eyes as a dictionary. The average length of an entry is 108 words.

Consider a typical example: Cawdrey’s and Cowell’s treatment of the word bail. Cawdrey gives two synonyms, while Cowell gives an etymology, two main definitions, two subentries (common bail and special bail), and an extended quotation from a treatise.

That’s emblematic of the differences to be found entry by entry – though the choice of headwords typically varies. One last point is worth making: half of the Cawdrey definition is entirely inaccurate: witness is in no sense a synonym of bail.

Not for a half-century would there be another significant monolingual dictionary: Thomas Blount’s Glossographia appeared in 1656. And this hard-word lexicon was by none other than a legal lexicographer – since Blount became equally well known for his Nomo-Lexicon, first published in 1670.

It was the legal lexicographers Cowell and Blount who, by quoting extensively from books and manuscripts, anticipated the great innovation of Samuel Reference JohnsonJohnson’s 1755 Dictionary – providing illustrative quotations, a technique carried forward and systematized in the Oxford English Dictionary. The most recent editions of Black’s Law Dictionary (Black’s) extend the technique by making the quotations not merely illustrative but also substantive and explanatory – again, a technique anticipated by Cowell and Blount.

In sum, legal lexicography gave English lexicography a major boost. Yet legal lexicographers have hardly been accorded their historical due. Indeed, only Thomas Blount receives any notice (Reference OsseltonOsselton 2009) in the two-volume, thousand-plus-page Oxford History of English Lexicography (Reference CowieCowie 2009). Even there, Blount’s law dictionary receives no mention.

Given the scant attention accorded to them, we might usefully trace the history of some notable legal lexicographers and their work. Six stand out: John Rastell, John Cowell, Thomas Blount, Giles Jacob, John Bouvier, and Henry Campbell Black.

26.3 The Challenges of Modern Legal Lexicography

There’s a great deal of naïveté about dictionaries. Some people think that legal definitions are immutable. They sometimes express discomfort that definitions have changed in recent editions of Black’s. In fact, it wasn’t particularly that the meanings of the words had changed – just that the accuracy of defining had. Although the methods of lexicography have been relatively stable over the years, computers have made things vastly easier in three ways.

First, we can now know with great reliability how old a term is in the English language. For Black’s, Fred Shapiro of Yale Law School has affixed a date to each headword so that users can easily see the earliest known use of each term. So under privilege, a word that dates from the 1100s, there are various types of evidentiary rules listed: accountant-client privilege (1956), attorney-client privilege (1934), bank-examination privilege (1990), doctor-patient privilege (1954), informant’s privilege (1962), journalist’s privilege (1970), marital privilege (1902), privilege against self-incrimination (1891), state-secrets privilege (1959), and tax-return privilege (1980). Ascertaining those dates was no simple matter. It took a great deal of ingenuity and knowledge for Shapiro to research and record this information. He has gone systematically through the dictionary to date almost every term. (See also Reference ShapiroShapiro 2018.)

Second, big data can tell us about the relative frequency of terms over time. For example, when deciding whether to put the main entry under exemplary damages or punitive damages, the Black’s editorial team created a Google n-gram to find out that (1) exemplary damages is the older term, dating from the 1700s, (2) punitive damages as a lexical item dates from the mid-1800s, (3) punitive damages became the predominant term in American English in the 1930s and in British English in the 1970s, and (4) today, in English worldwide, the phrase punitive damages is six times as common as exemplary damages. What an astonishing tool this is.

Third, the searchability of databases makes finding relevant texts, such as those that contain scholarly attempts to define terms, easier than ever. That’s been so since the early 1980s. Legal lexicographers had a headstart on English-language lexicographers because, by 1981, Westlaw and Lexis had created large corpora of fully searchable texts. In the 1980s, it was possible to know the relative frequencies of terms. In those days, there were arguments that conclusory wasn’t a “real” word, and that conclusionary or conclusional should appear instead. (See Reference GarnerGarner 1987, 135–136.) By the mid-1990s it was possible to trace conclusory in its modern legal sense to a 1923 New York opinion and to show that it predominated overwhelmingly in legal texts over the other two forms; by 1988, the term had already appeared in more than 21,000 published judicial opinions (see Garner 1995, 191–192).

But computers don’t replace old-fashioned book research. The Black’s team couldn’t conduct the kind of research necessary for their work without consulting hard-copy books. The eleventh edition of Black’s contains a bibliography of more than 1,000 treatises that are quoted in the main dictionary. These quotations had to be found the traditional way, by combing through the pages of reputable texts.

Consider the phrase international law, dating from 1786. The current entry has a traditional definition (“the legal system governing the relationships between countries”) followed by a more modern one (“the law of international relations, embracing not only countries but also such participants as international organizations and individuals”). Then follow nine synonymous phrases, which could be found only by examining dozens and dozens of texts on international law. And then there are two extended quotations from classic treatises: Jessup’s 1948 treatise called A Modern Law of Nations and the 2012 multivolume Max Planck Encyclopedia of Public International Law. Only by consulting dozens of treatises on the subject might one be satisfied that these two passages qualified as canonical treatments worthy of inclusion in Black’s.

Someone Googling the phrase would never come up with those two statements, or anything else nearly as quotable and citable. And it’s often difficult for a researcher looking at internet discussions at various websites to know the accuracy or authoritativeness of those sites.

Four big questions confront the writer of a modern law dictionary – and they’re pretty much the same questions faced by lexicographers of old, from Rastell to Jacob to Bouvier to Black. They are:

• To what extent should a law dictionary be a dictionary – as opposed to a legal encyclopedia? That is, to what extent should it merely define terms, rather than expansively discuss the law relating to those terms?

• To what extent is a law dictionary a work of original scholarship – rather than a compilation of definitions taken from judicial opinions?

• To what extent should we worry about the formalities of defining words – that is, about getting the lexicography right as well as getting the law right?

• To what extent can the modern lexicographer rely on the accuracy of predecessors?

A practicing lexicographer must answer those questions – sometimes ad hoc, from day to day and week to week. The answers largely explain why the seventh through to the twelfth editions of Black’s (1999–2024) look so different from earlier editions. Let’s take these questions one at a time.

26.4 Courts’ Reliance on Dictionaries

Courts use dictionaries in two ways. First, they sometimes need to ascertain the meaning of an unfamiliar word or phrase. Second, and more often, they need to verify the presumed meaning or meanings of a word or phrase. Native speakers of the language may know the words they use, but it can be helpful to see the semantic contours that expert definers have delineated. Frequently it’s only in litigation that the full accuracy of a definition can be tested.

Consider, for example, a hypothetical but illuminating case. A court must decide whether a statute referring to poultry applies to fighting cocks. The jurisdiction, which allows cockfighting, passed a 1978 statute granting a major tax exemption to those who raise poultry. A cockfighting operation has applied for the exemption, claiming that its birds qualify as poultry. The state disagrees. If the judges approach the question as textualists, they will want to discover the “fair meaning” of the word poultry. But that isn’t the only way of analyzing the problem. If the judges approach the question as so-called purposivists, they would want to know what the legislature was trying to achieve with the statute – perhaps without putting too much weight on the legislature’s choice of the word poultry. If the judges were consequentialists, they would be guided by their own views of good policy, putting very little weight on the precise wording.

So those with a textualist bent are most likely to consult and cite dictionaries. Interestingly, it’s easy to find definitions defining poultry as “domestic fowls collectively, such as chickens.” The five dictionaries in the court’s library – the American College Dictionary (1947), Funk & Wagnalls New College Standard Dictionary (1956), the Doubleday Dictionary for Home, School, and Office (1975), the Concise Oxford Dictionary (1999), and The Times English Dictionary and Thesaurus (2000) – are surprisingly uniform. From these authorities it would be possible to reason that (1) fighting cocks are raised in pens and not found in the wild – and are therefore “domestic”; (2) these definitions are broadly worded, and “collectively” connotes generality; and (3) fighting cocks are indeed literally chickens. So the cockfighting operation wins its tax exemption.

But is that result textually correct? Certainly not. The court’s five dictionaries missed the crucial element that poultry serves “as a source of eggs or meat” (Webster’s Third). If the judges had relied on Webster’s Second (1934), Webster’s Third (1961), the Random House (Reference Flexner1987), the Shorter Oxford (1993), Webster’s New World (2007), or the American Heritage (2011), they would have encountered the central idea that poultry refers to domesticated birds raised as a source of food.

So it can be good to look at many dictionaries and be selective. Or, as two commentators have put it, “a comparative weighing of dictionaries is often necessary” (Reference Scalia and GarnerScalia and Garner 2012, 417). The illustration just given shows why. If a court varies in its use of dictionaries, it will assuredly be criticized for “dictionary-shopping.” (See Reference AprillAprill 1998, 275; see also Reference Kirchmeier and ThummaKirchmeier and Thumma 2010, 77.) Such an accusation, though, may be premised on the false notion that a single dictionary is likely to be uniformly reliable in its definitions.

In recent decades, the Supreme Court of the United States has placed ever-greater weight on dictionary definitions. From 1800 to 1969, the Court invoked dictionaries in only 145 opinions; in the single decade from 1990 to 2000, the number surged to 239 opinions; the next decade came in at 225 (Reference Brudney and BaumBrudney and Baum 2013, 483, 495). By 2018, the Court was citing dictionaries in half its opinions (Reference LemleyLemley 2020, 1).

Three reasons are posited for the increased use of dictionaries: (1) the rise of textualism, especially as practiced by Justice Antonin Scalia, who influenced his colleagues and successors; (2) an increase in criminal-law cases, in which it is important to know the precise meaning of what conduct has been prohibited; and (3) the perception that reliance on dictionaries, as objective definers, counteracts the idea that judges are free-wheeling consequentialists engaging in judicial activism (Reference CalhounCalhoun 2014, 484).

Although one might expect the “conservative” justices to cite dictionaries more often than their “liberal” colleagues – the quotation marks signal the simplistic nature of these convenient labels – that has not been so (Reference Brudney and BaumBrudney and Baum 2015, 108). From 2005 to 2010, there was no meaningful statistical difference in the rate of citations in majority opinions (Reference Brudney and BaumBrudney and Baum 2015, 108).Footnote ¹⁴

One frequent criticism is that dictionaries encourage judges to engage in a kind of acontextual analysis (Reference EasterbrookEasterbrook 1994, 67; Reference RandolphRandolph 1994, 74). Even textualist judges can be acutely aware of the problem. Justice Scalia, for example, dissented in a case involving the meaning of the verb use.Footnote ¹⁵ A statute provided for the enhancement of a criminal sentence if the perpetrator of a drug crime “uses a firearm” in committing the offense. In the case at issue, the perpetrator had traded a gun for drugs. Does that constitute using a firearm? The majority said yes,Footnote ¹⁶ citing broad dictionary definitions of use. In dissent, Justice Scalia said no: using a firearm in this context denotes an act or threat of gun violence. He warned that dictionaries must be consulted bearing in mind that “the meaning of a word cannot be determined in isolation.”Footnote ¹⁷ His illustration was apt: “To use an instrumentality ordinarily means to use it for its intended purpose. When someone asks, ‘Do you use a cane?,’ he is not inquiring whether you have your grandfather’s silver-handled walking stick on display in the hall; he wants to know whether you walk with a cane.”Footnote ¹⁸

26.5 The Lexicography Rule: Every Lawyer a Lexicographer

Drafters of legal instruments – statutes, regulations, contracts, wills, and the like – commonly provide definitions of key terms. While drafting their documents, lawyers become ad hoc lexicographers. Textualists hold that their definitions must be “carefully followed” (Reference Scalia and GarnerScalia and Garner 2012, 225).

What is the origin of this doctrine? It can be dated back to the Late Roman Republic (ca 146 BCE to 31 BCE), when Roman jurists began to employ a technique known as grammatical interpretation. This early textualist method “concerned itself with a search for the meaning of the words, particularly with regard to the sense in which they were used by the writer” (Reference SchillerSchiller 1941, 749). Cicero, an adherent of this school of thought, claimed that a reader should frequently consider the drafter’s understanding of an expression – not just common usage – to arrive at the writer’s meaning (Reference SchillerSchiller 1941, 763). It takes only the slightest intention of this doctrine to say that if a drafter explicitly gives a definition for an expression, an interpreter should heed it.

In American law, the idea appeared as early as 1811, when the Virginia high court declared: “When, in the context of a will, the testator has explained his own meaning in the use of certain words, the court should take that as their guide, without resorting to lexicographers to determine what those words ought to signify in the abstract.”Footnote ¹⁹ By 1857, it was explicitly said that a will drafter could become “his own lexicographer.”Footnote ²⁰ Since the mid-nineteenth century, the doctrine has been extended to contracts, regulations, and statutes generally.

But as a fundamental tenet, the doctrine has become firmly rooted in the law of intellectual property, especially patents. At the beginning of the twentieth century, the United States Court of Appeals for the Seventh Circuit decided a case in which a patent drafter specified a peculiar definition for the word trestle – one that differed from those found in dictionaries. The court held that the drafter’s stipulation of an unconventional meaning should not deprive him of protection in his invention.Footnote ²¹

In 1967, the United States Court of Claims explained the doctrine’s rationale: “Often the invention is novel and words do not exist to describe it. The dictionary does not always keep abreast of the inventor. It cannot. Things are not made for the sake of words, but words for things. To overcome this lag, patent law allows the inventor to be his own lexicographer.”Footnote ²² Specifically, the innovation inherent in patents helps explain why neologisms would take hold and new meanings would need to be given to established expressions.

Yet the courts are wary of how the “lexicographer rule,”Footnote ²³ as it is sometimes called, might be strategically used to smuggle unexpected meanings into what might otherwise appear to be anodyne provisions. Judicial opinions will often contain cautions. As a matter of notice, the meaning must be clearly expressed in the written description.Footnote ²⁴ The definitions must be phrased “with reasonable clarity, deliberateness, and precision.”Footnote ²⁵

Counterintuitive definitions, though frequently discouraged in the literature on legal drafting,Footnote ²⁶ are said to be accepted by the courts. The most extreme expression of this idea is that a patent applicant, “acting as a lexicographer, may define ‘black’ as ‘white.’”Footnote ²⁷ Among the most notorious instances of this distortion is the perhaps apocryphal regulatory definition of milk as including all citrus fruits.

Not all the dangers inherent in departing wildly from ordinary meaning are self-evident. One of them is that the drafter will lapse into “using the word in its ordinary sense rather than in its defined sense, making nonsense of the inclusion of the extraordinary meaning” (Reference DickDick 1985, 77). Another is that a definition may not be fully considered in all the contexts in which it is used in a given legal instrument, so that “its inadequacies may not become apparent until the document has been executed” (Reference Doonan and MacfarlaneDoonan 1995, 85).

These are only some of the reasons why legal drafters are urged to be abstemious lexicographers.Footnote ²⁸ Lawyers are not, as a whole, skillful at defining. They violate the principle of substitutability. They use the definiendum within the definiens. They define terms that, oddly enough, are never used within the document. (They forget to use a term that, at the outset, they defined with the intention of using it.) And they wrench the crucial words from a given provision by placing them in a separate definitional section that is often wildly separated from the operative context – perhaps by fifty or one hundred pages. They are lexicographers with no training in lexicography.

Most of all, though, the lexicographer rule encourages legal drafters to forget that the Ordinary Meaning Rule (Reference Scalia and GarnerScalia and Garner 2012, § 6, 69) – the principle that words are to be understood in their everyday meanings – is an ally, not an enemy. They come to believe that even the most basic terms, such as conjunctions, must be defined. This definition, for example, appeared in a 2019 contract:

A client paid for this verbiage. Imagine what might happen when this drafter tries defining articles, prepositions, and be-verbs.

26.6 Conclusion

The law is a profession of words. Lawyers and judges are paid to do two things: speak and write. All other skills are subskills that merely precede the ultimate performance of speaking or writing. It is therefore unsurprising that lawyers and judges are more interested in dictionaries than, say, medical doctors, architects, and bankers.

In the words of United States Supreme Court Chief Justice Earl Warren, a court “does not review congressional enactments as a panel of grammarians; but neither do we regard ordinary principles of English prose as irrelevant […].”Footnote ²⁹ That has been steady throughout time. The words of US Supreme Court Justice Samuel Freeman Miller in 1873 would have been just as applicable in any age: the courts determine meaning “by the connection of the sentence, the meaning of the terms, and the rules of logic as well as grammar.”Footnote ³⁰ More recently, Justice Elena Kagan, also of the US Supreme Court, has written, “The question is […] simply: Is that the right and fair reading of the statute before us?”Footnote ³¹ Lexicographers seek to provide the needed guidance; lawyers and judges understandably ferret out and use what the lexicographers say. Both sides, at their best, pursue wisdom.