Digital corpus methods have become central to examining evidence in English language history. Since their advent in the 1990s, digital historical corpora have been an important way to bring particular linguistic features into transhistorical conversation with one another and to look at language change comparatively – they galvanized quantitative research in historical English linguistics. More recently, researchers have also pursued more varied ways to use corpora for qualitative investigation, opening the door to mixed-methods approaches and bringing balance to research tools for corpora. Daniela Landert’s innovative approach in Methods in Historical Corpus Pragmatics provides a valuable model for researchers wanting to use corpora for mixed-methods studies.

Methods in Corpus Pragmatics investigates Early Modern English documents and provides an important look at the ways in which they approach epistemic stance: the words and constructions that a text employs to express attitudes and to mark degrees of certainty about information and reliability of evidence. The study lies at the intersection of corpus methods and historical pragmatics, and it includes several case studies that investigate four different historical corpora and several genres of Early Modern texts.

Centrally, the examination targets and analyzes what Landert calls ‘high-density passages’ for indicating epistemic stance. Stance markers are not evenly distributed in discourse, it seems, and a reader who only studied them through concordance software or a Key-Word-In-Context (KWIC) application might not experience how the lexical markers for stance tend to cluster and work together to create different effects of intensity, subjectivity and positionality. Qualitative examination of the texts can show this more clearly: close reading of particular sections and passages where clustering takes place. The high-density passages show how writers use lexical and grammatical stance markers to function not as discrete entities but, rather, collectively: recruiting even words and constructions which might not typically be seen as stance markers for the stylistic effect. This is an important insight and goes some way to bridging linguistic methods for examining language with stylistic, rhetorical or literary analysis of fuller passages of text.

The monograph is structured in three parts and nine chapters. After the introduction, chapter 2 (pp. 13–38) presents a description of the field of historical corpus pragmatics. This summary sets out the development of historical pragmatics in the last few decades and the ways that the field has built upon and been enriched by digital corpora. The chapter describes the strengths of the field as well as its attendant challenges: scarcity of data, mark-up and formatting incompatibility between corpora, complications of trying to scale qualitative methods in order to apply them to larger datasets, and shifting relationships between linguistic form and function over time. Our pragmatic strategies for accomplishing pragmatic and communicative goals change over time, and therefore the mappings of form-to-function and function-to-form are unstable and inconsistent in historical data. Historical corpus pragmatics as a field must confront all of these concerns and find methodological strategies for accommodating them. This chapter also introduces and contextualizes the corpora which are used: the Corpus of Early English Dialogues 1560–1760 (CED), the Corpus of Early Modern English Medical Texts (EMEMT), the Lampeter Corpus (LC) and the Parsed Corpus of Early English Correspondence (PCEEC).

Chapter 3 (pp. 39–68) provides an introduction to the study of stance, setting out the relationship between evidentiality, the source of the information and epistemicity, i.e. the speaker’s certainty about the information and commitment to the truth of it. It also explains how researchers have categorized the types and usage of stance markers and how they have addressed the role of context.

The methodology for the study is set out in part II. Chapter 4 (pp. 71–118) describes how Landert compiled an inventory of lexical stance markers, putting together examples from previous research and filling out examples with searchable reference materials like the Oxford English Dictionary and the Historical Thesaurus. She also sets out criteria for classifying stance markers (e.g. expressive, evidential, attributed) and for assessing which are relevant for a particular study. Inventories of stance markers, it is clear, should be tailored to a particular analysis, focusing on the relevant features as they are established in a particular text, genre or period.

Chapter 5 (pp. 119–40) lays out the research method for retrieving the high-density passages that Landert pursues: passages which contain higher concentrations of strategies for marking epistemic stance. As an illustration, consider one of the passages which she presents and discusses (pp. 119–21):

In this passage from a medical text in which the author argues for the use of alkalis instead of acids, Landert signals the rich use of stance markers (marked with italics), which position the narrative in various respects. She points out how first-person stance verbs (I have heard it reported, I confess, I suppose, I know not, I doubt), modal verbs (may, should) and adverbs (usually, perhaps) combine with nouns like Confidence, Argument, Hypothesis and Truth, and with adjectives (sure, impossible, untrue), and with the use of conditional and concessive constructions. Further, she elaborates, acknowledging the context is essential: the words constitute part of an allegorical dialogue between personified chemical substances, which provides crucial backstory for the assertions of certainty, attribution and reliability. We see in this discussion how Landert’s layered approach to these stance markers illuminates the pragmatic richness of stance in context.

One advantage of the methodology is that it can be nimble: adjusting to a set of practices in a particular sample of texts. Sometimes pragmatic aspects like stance are marked in multiple ways and to varying degrees, and it is important to acknowledge how linguistic features can take on different pragmatic roles in different contexts. Corpora can sometimes be blunt instruments in this respect, but we see here how we can essentially train the corpus to focus more adeptly on what we are searching. The methodology has the difficulty, however, that it requires the researcher to cull data from different corpora with different formats and languages. As historical corpus research has evolved, different digital tools are often frozen in the format that seemed best at the time of their construction. Compiling data from multiple corpora, therefore, can present logistical problems for searching; for instance, the mark-up systems are often incompatible. Landert clearly had the skills and energy necessary to make this kind of comparative search possible, but her work-arounds herald the greater problems that corpus researchers will have moving forward as platforms and mark-up tools change. (I myself have a small collection of corpora and searchable texts saved on various kinds of disks and CDs, which are now effectively inaccessible to me. Thinking too hard about this triggers despair.)

Part III of the book is devoted to exploring the results of this empirical analysis. Chapter 6 (pp. 143–69) discusses the 42 high-density passages that Landert isolated in the corpora and examines their epistemic stance markers. One relevant finding is that the list of the stance markers collected from previous studies, which typically examined the words in isolation or in short context windows, does not present a full picture of the range of markers and strategies that these passages employ. I am particularly interested in what we might consider contingent stance marking, that is: features that are not forcibly stance markers, but serve to indicate epistemicity in context. Words like collect, esteem, fancy, persuasion and forsooth, to name a few, are among those that Landert observes functioning as lexical stance markers despite not being included in previous stance analysis. Finding these is one of the strengths of Landert’s careful close analysis. Chapter 7 (pp. 170–213) excels in this regard, focusing on the contextual factors and rhetorical tools that work together with stance markers. Indicating positionality, we see, is not an all or nothing prospect, but rather an interdependent network of features where subjectivity in one place colors the surrounding discourse. Stance, it appears, leaks into adjacent words and clauses.

Chapter 8 (pp. 214–33) focuses on two constructions, I saw and I tell (you). These meta-communicative expressions are verbal processes, and they sometimes take on the force of marking epistemic stance. Landert excluded these expressions from the inventories that she collected, but here she shows how I saw and I tell (you) work in context. Examining particular environments for the markers permits her to look in detail at their usage: the modal verbs connected with them, how their appearance in genres like trial proceedings and comedy plays influences their functions, and the relevance of speaker roles and activity types.

Chapter 9 (pp. 234–78) presents a final case study of certainty markers (e.g. truly, sure, certainly) in three different genres, comparing the presentation of certitude and reliability. Examining medical treatises, political pamphlets and trial proceedings shows variation in the writers’ expressed levels of confidence in the truth of a proposition, responding to contextual factors such as implicit or explicit assumptions about the uncertainty of the information presented. Examining the varying functions of the certainty markers (predictions, emphasizing stance, response, argumentation and alignment) in these three genres shows how they primarily express a commitment to the truth of a proposition, particularly in the face of a perceived uncertainty on the part of the speaker or author. This final chapter blends lexicogrammatical analysis of the markers with genre analysis and qualitative analysis of the passages, and thus it serves as a valuable culmination of the methodology and a discerning consideration of the stylistic and rhetorical construction of certainty.

Overall, Methods in Historical Corpus Pragmatics offers a convincing and well-conceived examination of stance marking in Early Modern English texts, probing the multiplex methods for marking stance and exploring the interdependent words, constructions and contexts which collectively construct textual orientation and subjectivity. The rich analysis and nuanced discussion testify to Landert’s deep expertise with Early Modern English documents and genres.

The monograph makes several valuable contributions to research, but the one that I found most compelling is the way that it conceptualizes, locates and analyzes the ‘high-density’ passages of stance. In determining how to isolate these passages in the corpora and in carefully analyzing how stance is built not just through particular words, but through discourse, the work develops a sophisticated mixed-methods approach to pragmaphilology. Stance, Landert demonstrates, is fundamentally linked to style and to cultural and textual contexts. Her insightful use of examples and passages illustrates how she gets into the weeds of close reading and rhetorical interpretation. She does not merely parse sentences, and she does not merely count instances; rather, she engages, deeply, with textual contexts and communicative strategies. The monograph also offers commentary on its methodology to demonstrate how corpus research can be qualitative and stylistic. In the last year or two, we have started to hear more about how generative AI computational tools present new possibilities for using embeddings in a neural network to reveal previously hidden relationships between words in a corpus, so I was even more struck when reading this book by the staying power of close reading as one of our strongest methodologies.