1.1 Academese: Obtuse or informational?

Many speakers of English share the view that the language of academic writing is peculiar, not only different from everyday speech, but also different from most other registers of English. These perceived differences are not neutral. Rather, a common stereotype of academic prose is that it is deliberately complex, and more concerned with impressing readers than communicating ideas – all making it needlessly difficult to understand.

These attitudes are reflected in the label academese, which is usually defined with negative connotations. For example, the Tameri Guide for Writers (www.tameri.com) defines academese as ‘an artificial form of communication commonly used in institutes of higher education designed to make small, irrelevant ideas appear important and original. Proficiency in academese is achieved when you begin inventing your own words and no one can understand what you are writing.’

In fact, the suffix –ese seems to have taken on this broader meaning when attached to other roots. Thus, according to the on-line site www.wiktionary.org, officialese is ‘the typical language of [written] official documents, legalistic and pompous’; legalese is ‘wording that resembles how a lawyer writes, especially such that is confusing to the layperson’; and bureaucratese is ‘any language containing many non-essential words intended to imply more importance or intelligence than actually present’.

The primary focus of these negative attitudes is on the use of rare and obscure words and phrases. The perception is that these words are often chosen to impress readers rather than efficiently convey new information. The portrayal of academic prose in similar terms has been around for some time, as evidenced by Orwell’s famous critique in 1946:

Here we see Orwell being especially critical of academic writing in the humanities (specifically art criticism and literary criticism). The enterprise of research in the humanities is focused on our everyday experience. Unlike scientists, humanities researchers are not discovering and documenting new natural phenomena and processes. Rather, their work is more interpretive, attempting to describe and understand the human experience. Like the author of Ecclesiastes, readers might believe that ‘there is nothing new under the sun’ when it comes to discussions of the human experience – and as a result, expect that there would be little need for new technical vocabulary and complex grammatical constructions to discuss these familiar ideas, relationships, and experiences. This seems to be Orwell’s underlying complaint: humanities researchers mostly discuss phenomena that are already familiar to us all, but they do it in a style that is indecipherable to non-specialist readers.

These stereotypes about academese in humanities writing continue right up to the present. For example, Pinker (2014: para 4) discusses the question of ‘Why academics stink at writing?’, and notes that:

Texts that seem to illustrate this prose style are unfortunately not hard to find. For example, the journal Philosophy and Literature sponsored a ‘Bad Writing Contest’ from 1995–1998, which celebrated ‘the most stylistically lamentable passages found in scholarly books and articles published in the last few years’ (http://denisdutton.com/bad_writing.htm). The contest lampooned academese simply by quoting the academic prose of famous literary scholars. For example, the winner of the 1998 contest was Judith Butler, ‘a Guggenheim Fellowship-winning professor of rhetoric and comparative literature at the University of California at Berkeley’. The first-prize sentence, singled out for its ‘anxiety-inducing obscurity’, reads as follows:

As noted earlier, the popular explanation for academese in humanities writing is that it represents a form of deliberate obscurity. Pinker (2014), however, proposes a slightly different explanation, based on the underlying goals of humanities versus science research writing. Pinker describes the primary goal of humanities writing as ‘self-presentation’ rather than the communication of information. This goal results in a ‘self-conscious style’, where the author’s primary concern is ‘to escape being convicted of philosophical naïveté about his own enterprise’. In contrast, science research writing is associated with a ‘classic style’ focused on the communication of information: ‘The writer can see something that the reader has not yet noticed, and he orients the reader so she can see for herself. The purpose of writing is presentation, and its motive is disinterested truth. […] The writer knows the truth before putting it into words; he is not using the occasion of writing to sort out what he thinks’ (Pinker 2014: para 11).

The linguistic characteristics of the ‘classic’ style associated with science writing are dramatically different from the ‘self-conscious’ style of humanities writing. Science prose can be boring, relying on simple syntactic constructions, as in:

Although there are often technical terms in this style, those terms usually refer to physical entities rather than abstract concepts. The overall stylistic effect is to emphasize the direct communication of information (even if that information is still unclear to non-specialist readers because of the technical vocabulary).

Our goals in the present book are not to enter into the debate about ‘good’ and ‘bad’ academic writing, or the motivations for the ‘academese’ found in many academic texts. But we are interested in the linguistic characteristics of academic writing, including the differences between humanities writing and science writing. And we are especially interested in the ways in which academic writing has changed historically in its grammatical style.

Given the striking linguistic differences between present-day humanities writing versus science writing, readers probably will not be surprised to learn that academic writing styles have undergone major linguistic change over the past few centuries. In particular, it would be easy to assume that academic writing has become increasingly complex over time, and that humanities writing has been especially influenced by these changes, resulting in the grammatically-elaborated styles of academese illustrated in Text Sample 1.1. At the same time, it might be assumed that science prose has somehow resisted these historical changes, and thus retained the grammatically-simpler styles of earlier centuries.

It turns out that those expectations are wrong in nearly every way. In the following chapters, we show that:

• – There are different types of grammatical complexity.

• – Complexity in humanities writing is associated with structural elaboration.

• – Humanities academic writing has changed little over the last few centuries in its structurally elaborated grammatical style.

• – Complexity in science writing is associated with structural compression, not structural elaboration.

• – Science writing has undergone extreme historical change related to the use of these structural compression features.

In the next section, we provide more detailed discussion of the grammatical features associated with academic writing, and the striking grammatical differences between humanities and science writing. This is followed by sections that set the stage for our analyses: previous research on grammatical change, and the discourse factors that influence change. In contrast to most previous research – which focuses on spoken interaction as the breeding ground for linguistic innovations – we argue here that historical change can occur in writing as well as speech. The following chapters present the results of empirical corpus analyses, showing that this theoretical possibility has indeed occurred. In fact, the corpus analyses in following chapters show that academic science writing has been the locus of some of the most dramatic grammatical changes that have occurred in English over the past three centuries.

1.2 Describing the grammar of academic writing

Researchers in the humanities would argue that they do not simply document human experiences that are familiar to us all; rather, they are constantly offering new interpretations of those experiences, and as a result, coining new terms that reflect those new interpretations. Thus, although these researchers discuss familiar situations and experiences, they do so in highly technical ways that are not easily understood by the non-expert. For example, consider the following excerpt from a literary criticism article:

One of the most salient characteristics of this text is the highly specialized vocabulary. Abstract technical terms are common in this passage, such as interiority and fetishisms. In addition, relatively common words like things, objects, contours, interior, depth, and life are used together with technical terms to produce complex phrases referring to highly abstract concepts, such as the peculiar resonance of things, object-filled interiors, the contours of interiority, nostalgia for a more pristine and private form of psychological depth, the public empire of things, and a paradoxically fetishistic preoccupation with the traces and tokens of inner life. Taken together, these abstract terms and complex phrases make it difficult for the non-specialist reader to understand the content of the literary analysis presented in the article.

Research writing in the sciences is probably even more difficult for the non-expert to understand, but its linguistic style is less often singled out for criticism. In part, this is because there is less of a mismatch between our expectations and the discourse style. The general goal of science research is to discover new information about the natural world, identifying new phenomena and explaining previously identified phenomena and patterns. Scientists require new words to refer to these previously undocumented phenomena and processes, resulting in prose that can be almost unintelligible to the non-expert. However, since the non-specialist reader has no prior experience with or existing knowledge of those phenomena and processes, we have little expectation that we should fully understand these texts. The following passage from a biochemistry research article illustrates:

Similar to Text Sample 1.3, the most obvious characteristic of this biochemistry article is its technical vocabulary. Most readers of the present book have never encountered the verb to phosphorylate, or nouns like kinases, vimentin, prometaphase, metaphase, and phosphorylation. Since we have no idea what these terms refer to, the entire passage has little meaning for us. Thus, the vocabulary of science research writing is similarly technical to the vocabulary of humanities prose, supporting the stereotype that all academic writing is complex and hard to understand.

It is much more difficult to notice the typical grammatical structures used in these academic texts. However, when considered at the grammatical level, we discover that there are important linguistic differences among the various disciplines of academic prose; and in particular, the language of science research writing is quite different from the language of humanities prose. Such considerations lead in turn to other surprising findings that contradict previous claims and assumptions about academic writing. In particular, the corpus-based analyses presented in the following chapters challenge four major stereotypes about academic prose:

1. 1. all kinds of academic prose are essentially the same

2. 2. academic prose employs complex and elaborated grammar

3. 3. academic prose is maximally explicit in meaning

4. 4. academic prose is conservative and resistant to linguistic change

Along the way, we challenge two basic theoretical assumptions that have been widely adopted in previous linguistic research:

1. 1. grammatical complexity is equivalent to structural elaboration, realized especially through the increased use of dependent clauses

2. 2. grammatical changes are initiated in speech; grammatical innovations do not occur in writing

In the following sections, we briefly introduce each of these widely held beliefs and begin to challenge these assumptions.

1.3 Grammatical change in English: Below the radar?

Historical change in English is an important topic treated in almost all introductory textbooks on linguistics and sociolinguistics. For the most part, though, these treatments focus on changes in pronunciation/phonology or lexical change. For example, several books document phonological changes in English like the loss of initial [h] in ‘which’, ‘whether’, and ‘whine’, or lexical changes like chesterfield > couch and serviette > napkin (see, e.g., Holmes Reference Holmes1992; Chambers Reference Schneider, Chambers, Trudgill and Schilling-Estes2002). In many cases, these changes result in present-day patterns of regional or social dialect variation. Examples of this type include the loss or retention of post-vocalic [r], the merger of [t, d] with [Ɵ, ð], and the vowel shifts found in American northern cities (see, e.g., Wolfram and Schilling-Estes Reference Wolfram and Schilling-Estes2005).

Grammatical change is generally disregarded in introductory linguistics textbooks. The few books that deal with this topic at all tend to focus on morphological change, such as the regularization of verb paradigms for irregular verbs (e.g., dreamt → dreamed), or vice versa (sneaked→ snuck). Some books also discuss patterns of morphological change (or resistance to change) that have resulted in present-day dialect differences (see, e.g., Wolfram and Schilling-Estes Reference Wolfram and Schilling-Estes2005: 85–92; Finegan Reference Finegan2008). Such changes include the loss of third person singular verb inflection (he runs → he run) and the development of second person plural pronouns in many regional dialects of English (e.g., y’all, youse, yoins).

In this regard, the coverage in linguistics textbooks is typical of linguistics research generally, which places ‘much more emphasis on phonological and lexical phenomena than on grammatical ones’ (Leech et al. Reference Leech, Hundt, Mair and Smith2009: 7). One explanation for this fact is that we just fail to notice grammatical change: that ‘we are considerably less perceptive of it than of other kinds of linguistic change’ (Strang Reference Strang1970: 60, cited in Leech et al. Reference Leech, Hundt, Mair and Smith2009: 7). Leech et al. go on to discuss the difficulty in noticing grammatical changes. One obvious factor is the nature of the phenomena. When the phonetics of a word begins to change, we notice that some people pronounce the word differently from others. When a new word comes into the language, we notice that some people are using the word and others not. From a synchronic perspective, such changes usually manifest themselves as social patterns of dialect variation; consequently, we readily notice some groups of speakers using pronunciations and words that are not used by other groups of speakers. Such patterns of variation can evolve rapidly, so that there are often dramatic phonological and lexical differences from one generation of speakers to the next (see, e.g., Labov Reference Labov1994, Reference Labov2001, Reference Labov2010).

In contrast, grammatical change is usually more abstract, involving the use of different phrase types and different clause types, and changes in the ways in which those structures are configured relative to one another. Probably an even more important consideration is that grammatical change proceeds much more slowly than lexical/phonological change, and it is less clearly associated with dialect variation.

The grammatical changes that have been most often noticed are typically lexico-grammatical: changes in the grammatical function of specific words. Prescriptive grammarians often target specific lexico-grammatical features associated with the general ‘decay’ of written English. Most of these pundits focus on present-day use and perceived misuse, but the targeted features are often specific words undergoing grammatical change. Examples include variation in the past participle form for irregular verbs (e.g., dove vs. dived, learnt vs. learned, lit vs. lighted), preposition choice (e.g., different from vs. different than), less versus fewer with count nouns, like functioning as a subordinator or a preposition, the conversion of adjectives to adverbs (e.g., real good; go fast/slow), and the choice between that, who, and whom as relative pronouns with an animate head noun.

Similarly, sociolinguistic studies of grammatical change usually focus on developments that affect specific words. For example, numerous studies have documented the recent grammatical change associated with features like the new quotative verbs (e.g., go, be like, be all), the discourse marker like, and so used as an intensifier modifying an adjective (see the survey of research in Tagliamonte Reference Tagliamonte2006). Changes of this type affect only a few specific words, they occur in a relatively short time span, and they are usually associated with strong patterns of social dialect variation; as a result, such changes are often noticed and described by professional linguists as well as the wider public.

In contrast, other grammatical changes are general structural shifts, influencing an entire construction type or grammatical system. For example, over the last few centuries there have been general changes in the use of modal verbs (e.g., Leech Reference Leech, Facchinetti, Palmer and Krug2003, Reference 266Leech2011; Millar Reference Millar2009), verb phrases with marked aspect (e.g., Smith & Rayson Reference Smith and Rayson2007; Nesselhauf Reference Nesselhauf2007; Mair and Hundt Reference Mair and Hundt1995; Smith Reference Smith, Peters, Collins and Smith2002; Hundt Reference Hundt2004; Mair Reference Mair2006), and complementation patterns (e.g., Mair Reference Mair and Hundt1995, Reference Mair2002; Rudanko Reference Rudanko2010). Such changes have a lexical aspect, in that they gradually work their way through the lexicon. However, these are not innovations restricted to a few individual words. Further, in contrast to the restricted lexico-grammatical changes discussed in the last paragraph, these general structural changes occur gradually over the course of several centuries, and they typically do not serve to distinguish among social dialects during their period of transition. As a result, such historical developments have attracted much less attention than short-term, socially-stigmatized grammatical changes in the use of particular words.

For example, progressive aspect has increased strongly in use over the past three centuries, developing new meanings and functions which have gradually occurred with an ever increasing set of lexical verbs (see Hundt Reference Hundt2004; Mair Reference Mair2006; Leech et al. Reference Leech, Hundt, Mair and Smith2009). If we compare the end points of this historical change, the differences are as dramatic as the recent development of new quotative verbs like go and BE all. However, historical change in the use of progressive aspect has been much more gradual, occurring over a period of centuries rather than a period of a few years. And there is no particular verb that we can point to as the target of this change; rather, the use of progressive aspect has been generally increasing for an ever expanding set of verbs.

These two characteristics – long-term change (vs. rapid change) and change influencing a grammatical system (vs. change that targets a few specific words) – are typical of many grammatical developments in English. Such patterns of change are not salient to speakers of a language at any given point in time, and they are not obviously associated with synchronic patterns of dialect variation. As a result, linguistic change of this type has generally been disregarded in comparison to lexical and phonological changes. However, several recent books on the history of the English language include descriptions of such grammatical change. In the following section, we survey some of the major features discussed in those treatments.

1.4 Two types of grammatical change

A long-term perspective on the history of English makes it clear that there have been numerous important grammatical changes in English over the course of the last 1,000 years. These include the change to a fixed SVO word order, the loss of most inflectional morphology (especially case suffixes), the increase in the range of function words (including prepositions, auxiliary verbs, and infinitive marker to), and more recently the introduction of the dummy auxiliary DO (see Rissanen Reference Rissanen and Lass1999; van Gelderen Reference van Gelderen2006).

However, grammatical changes over the last 300 years – the period of Modern English – have been much less dramatic. There have been numerous interesting developments, but they differ in kind from the earlier dramatic changes. Recent historical developments include many of the features that have been discussed by prescriptivists, such as changing preferences for irregular versus regular verb forms, the choice of fewer versus less, and the decline in use for the relative pronoun whom. However, there have also been changing patterns of use for numerous other grammatical features, documented in survey articles by Barber (Reference Barber1964: 130–144), van Gelderen (Reference van Gelderen2006: 171–174, 214–220), and Fennell (Reference Fennell2001: 144–145, 173–174), as well as in the more extended treatments by Denison (Reference Denison and Romaine1998), Mair (Reference Mair2006), and Leech et al. (Reference Leech, Hundt, Mair and Smith2009). Some of the more important changes are:

• – development of auxiliary-like uses for want to → wanna

• – increase in the use of the semi-modals be going to (gonna) and have got to (gotta)

• – decrease in use of the core modal verbs

• – increasing use of analytical rather than synthetic comparison

• – increasing use of the s-genitive, especially with non-human nouns

• – decreasing use of the of-genitive

• – increasing use of the ‘mandative’ subjunctive

• – decrease in the use of shall as a future marker

• – increasing use of progressive constructions

• – development of the progressive-passive

• – development of the get passive

• – overall decreasing use of the be passive

• – increasing tendency to omit the relative pronoun that

• – increase in the number and type of multi-word verbs

• – changing complementation patterns with non-finite clauses

These linguistic developments illustrate two major types of grammatical change:

1. a) the emergence and development of new grammatical features (including new grammatical uses of an individual word); and

2. b) shifts in the typical use and functions of core grammatical features

The first type of grammatical change has been studied mostly under the rubric of ‘grammaticalization’, which focuses on the way in which content words evolve over time to be used as grammatical function words. Examples from the list mentioned earlier include the use of have to and got to as semi-modals, wanna with modal auxiliary functions, and get as an auxiliary verb in passive constructions. Other examples include the use of well as a discourse marker, pretty as a hedge or intensifier, and sequences like in spite of, with regard to, and because of used as complex prepositions (see, e.g., Krug Reference Krug2000; Tagliamonte Reference Tagliamonte, Lindquist and Mair2004; Hopper & Traugott Reference Traugott and Hickey2003; Lindquist and Mair Reference Lindquist and Mair2004; Nevalainen Reference Nevalainen, Lindquist and Mair2004; Hoffmann Reference Hoffmann2005).

The new quotative verbs illustrate a related kind of grammatical change. Some of these forms – like go – have evolved from a lexical verb to be used in the more grammaticalized function of marking reported speech and thought. The quotative be all also seems to have evolved from a lexical (adjectival) use to this more grammaticalized function. However, it could be argued that be like evolved in the opposite direction, from a function word (a preposition or subordinator) to a more lexicalized grammatical function as a quotative marker. Regardless, these are all examples of grammatical innovation that result in particular words changing to be used with new grammatical functions.

In contrast, the second type of grammatical change is completely different in kind: it focuses on the extent to which a grammatical feature is used rather than the development of new grammatical classes. This type of grammatical change affects the overall frequency of a feature and the co-occurrence of the grammatical feature with an ever increasing set of words; these changes are also often associated with an increase in the meanings and functions expressed by the grammatical feature. Examples of this type from the above list include the increasing use of the s-genitive, progressive verbs, and multi-word verbs. At the same time, other features have declined in use and functionality over the past few centuries (e.g., modals and passive voice verbs).

It turns out that most grammatical change in English over the past 300 years has been quantitative rather than categorical. Even grammatical innovations, like the development of semi-modals and the get-passive, have gradually continued to increase in frequency and functionality over this period. Thus, a quantitative perspective on grammatical change has become increasingly important in recent research:

Most of the diachronic changes that we document in this book are of this type, relating to the frequency of occurrence and the functionality of grammatical features, rather than the development of completely new grammatical categories. It turns out that grammatical change of this type has been especially prevalent in academic writing, unlike the lexico-grammatical innovations typical of speech.

1.5 The locus of historical change in English: Speech or writing?

It will be clear from the preceding sections that we are interested primarily in grammatical innovations that have occurred in the discourse of academic writing. For many historical linguists, this focus is a non-starter. That is, most previous research on grammatical change has focused on spoken discourse, and it has often overtly excluded the possibility of grammatical change in writing.

The view that language change occurs in speech – and not in writing – can be attributed to several sources. At the most basic level, this view reflects the widespread belief in linguistics that spoken language is ‘real’, while written language is a derivative artifact (see discussion in Biber Reference Biber1988: 5–7). In the early American linguistic tradition, Sapir and Bloomfield both held this position:

And numerous other linguists in more recent decades have reiterated this position:

Nearly all introductory textbooks on linguistics continue to reflect this bias in their reference to ‘speakers’ and ‘speech communities’, with little or no acknowledgement of ‘writers’ or ‘writing communities’.

Historical linguists who study recent change in English have similarly been preoccupied with the characteristics of speech, describing the ways in which English has become increasingly colloquial and informal in recent centuries. The linguistic features associated with these changes originated in speech, and they have gradually been adopted in writing. Thus, to the extent that historical researchers have considered change in writing, they have focused mostly on the increasing use of colloquial features borrowed from speech, such as first person pronouns, contractions, and semi-modals. This historical trend towards a more colloquial style has been documented by numerous discourse analysts and corpus linguists, and is referred to as the ‘drift’ of written registers towards more ‘oral’ styles (Biber and Finegan Reference Biber and Finegan1989a), ‘informalization’ (Fairclough Reference Fairclough1992), and ‘colloquialization’ (Hundt and Mair Reference Hundt and Mair1999; Mair Reference Mair2006; Leech et al. Reference Leech, Hundt, Mair and Smith2009). For example, Mair (Reference Mair2006) describes this process as follows:

Earlier investigations of these patterns suggested that they represent a general historical trend in English. However, subsequent investigations show that there are important differences among written registers. For example, Biber and Finegan (Reference Biber, Finegan, Nevalainen and Kahlas-Tarkka1997/2001) show that written registers like fictional novels and personal letters have been strongly influenced by the drift towards more colloquial linguistic styles, but written academic registers (especially science research articles and medical research articles) have not participated in that trend. Hundt and Mair (Reference Hundt and Mair1999) also note this difference, distinguishing between ‘agile’ written registers (e.g., newspaper prose) that are receptive to these changes, and ‘up-tight’ written registers (e.g., academic prose) that resist historical change towards the increased use of colloquial features.

A complementary approach to historical change has focused on the ways in which grammatical innovations emerge in language use, carried out under the rubrics of ‘grammaticalization’, ‘usage-based approaches’, ‘emergent grammar’, and the general study of ‘frequency effects’ in language use. These approaches share a focus on grammatical use in natural discourse, describing how new grammatical constructions and/or functions emerge from natural communicative situations.

Most usage-based studies of grammatical change focus on spoken interaction, using the term ‘speaker’ as a cover term for the addressor or producer of discourse, and often explicitly framing the discussion in terms of ‘utterances’ and conversational ‘interaction’.Footnote ¹ Written discourse has generally received little attention, with most studies simply disregarding the possibility that grammatical innovations might develop in natural written communication.

For example, an exclusive focus on spoken interaction is front-and-center in the edited book on Interaction and Grammar by Ochs, Schegloff, and Thompson (Reference Ochs, Schegloff and Thompson1996), which approaches grammar as part of the social practices associated with conversational interaction. The studies in this book analyze transcripts of conversations, showing how grammar emerges with special functions in spoken interactional contexts. A similar implicit emphasis on spoken interaction can be observed in so-called usage-based models of grammar (see, e.g., Langacker Reference Langacker1987; Kemmer & Barlow Reference Kemmer, Barlow, Barlow and Kemmer2000). While these descriptions tend to be based on intuitive notions of spoken interaction, rather than direct analysis of conversational texts, they still implicitly frame the discussion relative to spoken discourse; for example, the ‘usage events’ that form the foundation of the usage-based model are ‘utterances’ produced by ‘speakers’ (Langacker 2000: 9).

Similarly, in describing the motivations and enabling factors of grammaticalization, Hopper and Traugott (Reference Traugott and Hickey2003: 71) note that previous research has focused on ‘the role of speakers and hearers negotiating meaning in communicative situations’. Bybee and Hopper (Reference Bybee and Hopper2001) are also typical in their focus on spoken interaction, noting that:

Croft (Reference Croft2000) is especially emphatic in arguing that language change occurs in utterances produced by speakers in conversational interaction; for example:

Given this background, it is not surprising that many studies of grammaticalization have focused on grammatical features that are common in conversation but rarely used in writing, such as the English semi-modals (e.g., have to, got to), discourse markers (e.g., well, like), intensifiers (e.g., so), and quotative verbs (e.g., go, be all).

As noted above, most scholars simply disregard the possibility that grammar might also emerge in written use. One of the few exceptions is Croft (Reference Croft2000), who directly considers the possibility but then strongly argues against it:

In this passage, Croft is specifically discussing grammatical innovation from a typological perspective: can a new type of grammatical construction emerge in a written language in comparison to the types of constructions that already exist in the world’s spoken languages? However, his basic position is in line with the generally accepted view of most historical linguists: that grammatical innovations occur in speech, not writing; and that any grammatical developments in writing are derivatives from speech.

Of course, historical linguists do not have access to speech samples from earlier historical periods, and thus they are forced to rely on the analysis of written texts. It might be supposed that this methodological requirement would have forced linguists to consider the possibility of grammatical change originating in writing. But that has not occurred. In fact, the methodological focus of historical research has been on ways to filter out the assumed ‘obscuring’ effect of writing, in the attempt to uncover the ‘true’ patterns of language use that occurred in speech from earlier periods.

Schneider (Reference Schneider, Chambers, Trudgill and Schilling-Estes2002) has written one of the most thorough treatments of this methodological issue, under the promising title ‘Investigating variation and change in written documents’. This title might lead the reader to expect discussion of the importance of grammatical change in writing. However, the very first sentence of the article immediately quashes that possibility, asserting that speech is primary and writing merely ‘a secondary coding’ (Schneider Reference Schneider, Chambers, Trudgill and Schilling-Estes2002: 67). Schneider goes on to discuss the implications of this view for historical research:

In sum, Schneider argues that historical linguists have no choice but to analyze written documents. However, they should have no expectation of uncovering grammatical changes that emerge in writing. Rather, the focus is exclusively on speech (representing the real ‘language system’), and the methodological challenge is therefore to uncover the patterns of use that existed in speech, based (unfortunately) on the analysis of written texts.

Culpeper and Kytö (Reference Culpeper and Kytö2010) directly address this methodological challenge by carrying out detailed linguistic analyses of spoken dialogue as represented in speech-based registers from Early Modern English (e.g., legal trial proceedings, witness depositions, drama). Other researchers (e.g., Nevalainen and Raumolin-Brunberg Reference Nevalainen and Raumolin-Brunberg1996; Denison Reference Denison and Romaine1998) have collected texts from informal/colloquial registers, like personal letters or diaries, in the hope that these will approximate the underlying linguistic system found in speech from earlier historical periods. These studies have been highly important for the wealth of detailed information that they provide about the typical linguistic patterns of use in earlier periods. However, they also reflect the dominant view that the primary inquiry of interest in historical linguistics concerns the patterns of change that occurred in speech. Written records allow us to infer what those changes have been, but they are assumed to be of no interest in themselves. And in particular, the language of informational written prose is considered to be far removed from the kinds of discourse that historical linguists should be analyzing to uncover genuine historical change.Footnote ²

In the present book, we take a fundamentally different position, arguing that the communicative demands of any register – whether spoken or written – have the potential to facilitate the emergence of grammatical uses (and possibly new constructions) associated with those communicative needs. The first author has been advocating this position for the last 25 years; for example:

Biber and Conrad (Reference Biber and Conrad2009: 6) further develop this argument, claiming that ‘linguistic features are always functional when considered from a register perspective’; that ‘linguistic features tend to occur in a register because they are particularly well suited to the purposes and situational context of the register’. These functionally-motivated patterns of linguistic variation characterize present-day spoken registers (Chapter 4) and written registers (Chapter 5), as well as historical change in writing (Chapter 6).

Fox (Reference Fox2007: 299) has also advocated investigation of this possibility, stating that: ‘The relationship of written language to language-in-interaction is complex and worthy of independent study’. Thus, Fox argues that written discourse is produced under completely different circumstances from spoken discourse, and that these characteristics might have direct grammatical consequences:

The present book explores this possibility, based on the underlying assumption that the communicative circumstances of any register will provide the context for functionally-motivated grammatical change. In particular, we focus on academic research writing as a register that differs in almost every way from face-to-face conversation:

• – written rather than spoken

• – monologic rather than interactive and co-constructed

• – requiring specialized, professional background knowledge, but no assumption of personal background knowledge

• – slowly produced and carefully revised and edited

If grammatical change is influenced by the pressures of the communicative situation, there is every reason to expect that the grammatical devices and functions emerging in written academic writing will be strikingly different from those that have emerged from conversational interactions.

The analyses in the following chapters show convincingly that historical written texts should not be treated as a ‘filter’ that obscures true historical change. Rather, extensive grammatical change has occurred in writing. These patterns of change have affected different grammatical characteristics from the colloquial features described in most previous research. However, the grammatical innovations of written discourse cannot be discounted as mere ‘iterations and embeddings’. In particular, focusing on noun phrase constructions in English, the following chapters document extensions in the range of grammatical variants, the range of lexical associations, and the range of grammatical/semantic functions – in addition to dramatic quantitative changes in the frequency of use.

1.6 Overview of the present book

The present book challenges previous assumptions about grammatical complexity, historical linguistic change, and academic writing, documenting the ways in which academic writing has been linguistically innovative over the past two centuries. The primary focus of the book is the description of phrasal complexity features and the associated phrasal discourse style that is typical of present-day science research writing. From a synchronic perspective, the book documents how academic written registers are unlike other registers – spoken or written – in their reliance on phrasal complexity features rather than a clausal style of discourse. From a diachronic perspective, the book documents linguistic innovations that have emerged in writing (especially science research writing) but have generally not been adopted in speech.

We base our descriptions on large-scale corpus analysis that allows us to track long-term change in the grammatical structure and discourse style of academic writing, as compared to a range of other written and spoken registers. Chapter 2 introduces the corpus-based approach to linguistic analysis, as well as the specific corpora and analytical methods used for our study. It begins with a discussion of recent trends and issues in diachronic corpus-based research. The chapter then introduces the corpora that serve as the basis for our analyses; these span 300 years and represent a range of registers (e.g., everyday conversation, newspaper writing, fiction, academic writing). The academic writing corpus is designed to represent several sub-registers, including humanities prose, social science articles, and natural science articles.

Chapter 2 also outlines the research methods for our analyses, including corpus annotation, interactive and automatic analysis techniques, and quantitative research designs. Finally, Chapter 2 introduces and exemplifies the grammatical features discussed in the book, including the full set of clausal and phrasal complexity features.

Chapters 3 through 6 present the major findings of our research. Chapters 3 and 4 present quantitative corpus findings, while Chapters 5 and 6 are more interpretive, presenting the results of detailed linguistic analyses. These chapters are sequenced as follows: Chapter 3 describes synchronic patterns of register variation; Chapter 4 presents the results of similar analyses carried out from a diachronic perspective; Chapter 5 presents detailed linguistic analyses to support the claim that these historical developments are not ‘merely’ stylistic changes in frequency; and finally, in Chapter 6, we turn to the question of explicitness, showing how phrasal complexity features are less explicit in meaning than alternative clausal features.

Chapter 3 provides a synchronic description of the grammatical discourse style of academic writing, focusing especially on features associated with grammatical complexity and elaboration, but also describing the use of common, core grammatical features. A major goal of this chapter is to show that traditional operational definitions of ‘complexity’ – in terms of clausal embedding – fail to capture the actual complexities of academic writing, which are phrasal in nature. Through a series of case studies, we undertake synchronic comparisons between academic writing and other spoken and written registers to demonstrate that this phrasal discourse style is highly distinctive – occurring only in written registers with highly informational purposes and specialized audiences. Thus, this chapter documents the distinctive grammatical characteristics of modern academic research writing by comparing it to conversation, textbooks, classroom teaching, fiction, and newspapers. We conclude Chapter 3 with a case study that considers variation within modern day academic research writing. Chapter 3 demonstrates that academic writing – and especially science research writing – can be regarded as an outlier: a register unlike nearly all other spoken and written registers.

Chapter 4 then turns to a diachronic analysis of this same range of registers, to document the evolutionary path that academic research writing has taken over the past 300 years. We investigate the use of colloquial features as well as complexity features, and we consider both clausal as well as phrasal complexity features. The results show that academic writing has been resistant to change in the use of colloquial features, but it has actually been the locus of grammatical change in the use of phrasal complexity features. Comparison of sub-registers within academic writing show that specialist science research articles have been especially important for these grammatical innovations.

Chapter 5 presents detailed linguistic analyses to support the claim that these historical developments are not ‘merely’ stylistic changes in frequency. Rather, this chapter documents major extensions in the lexical associations, meanings, and discourse functions of phrasal complexity features. Chapter 6 continues our functional analysis of the phrasal discourse style of academic writing, turning to an unintended consequence of these grammatical innovations: the loss of explicitness. Through detailed analyses of particular phrasal devices, we discuss how phrasal noun modifiers compress information into dense constructions that are considerably less explicit in meaning than alternative clausal features.

Taken together, the analyses presented in Chapters 3 through 6 directly challenge many stereotypes and assumptions about grammatical complexity, historical change, and academic writing. We summarize the overall progression of the book in Table 1.4.

The final chapter, Chapter 7, summarizes the main themes of the book, leading into a discussion of applied implications. First, the chapter revisits these basic assumptions about complexity, linguistic change, and academic writing, highlighting how the analyses presented in Chapters 3–6 counter previous stereotypes. We then turn to the implications that result from our corpus-based findings. We discuss implications for theories of grammatical complexity, arguing that phrasal modifiers and structural compression are equally important to, but fundamentally different from, clausal modifiers and structural elaboration. We also consider implications for linguistic change, arguing that grammatical change can emerge in writing, and that in fact certain types of linguistic change must be considered within the context of written language. Finally, we explore applied implications related to the study of language development and the teaching of academic reading and writing (English for Academic Purposes).