Mysteries of English Grammar

A Guide to Complexities of the English Language

Andreea S. Calude, Laurie Bauer

• Publisher: Routledge
• Year: 2021
• ISBN:
  9781000451726
• Edition: 1
• Topic: Lingue e linguistica
• Subtopic: Grammatica e punteggiatura

Despite a history of hundreds of years of research analysing aspects of English grammar, there are still open problems which continue to baffle language researchers today. Such 'grammar mysteries' arise for a number of reasons: because the language is changing; because different speakers of the language adhere to distinct norms and thus introduce and maintain variation in the system; because there are differences between the grammar of spoken and written English. This book illuminates some of the complexities of the subject, the areas where new discoveries await and why it matters.

Through a series of accessible and engaging case studies on various aspects of grammar, from multiple negation to possession, the authors present grammar as an intellectual challenge. This book brings out into the open questions about language usage to which we still do not have good answers in a bid to make variation overt and to revel in the mystery of the English language.

Both aimed at the interested general reader and the beginning student of English language and linguistics, this is a fresh take on grammar.

1 Introduction

Laurie and Andreea

DOI: 10.4324/9781003148999-1

Setting the scene

English is currently the most described language in the world. We say this not only because of the number of very detailed grammars of English such as Jespersen (1909–1949), Quirk et al. (1985) or Huddleston & Pullum (2002), and not only because we have been writing grammars of English for a very long time, one of the earliest works being Bullokar’s Pamphlet for Grammar in 1586 (Linn 2006, p. 74). While both these factors are clearly important, it is the range of constructions which are part of English that have been subject to detailed consideration by generations of grammarians and linguists, either trying to provide a better description of English, or trying to prove some theoretical point about the way grammar works. The descriptions do not always agree (for various reasons, some of which we will consider here), but they are available for consultation, and they provide a very detailed description of more of the grammar of English than has been attempted for any other language. It might thus seem that yet another book about English grammar would be a waste of time. Surely, it has all been done, and the job of English grammarians has now been rendered redundant.

In this book, we argue that this is far from being the case. Researchers who work in the field of English grammar are still making new discoveries, finding new regularities and providing new insights. There are many reasons why we might not know precisely what is going on in grammar, not only in English, but in other languages, too.

Variation in language

First, we should note that English is not a homogeneous language. There are many varieties of English, where ‘variety’ is a technical term that encompasses dialects, styles, genres, even patterns which may be indicative of the individual speaker or writer. The examples given in (1)–(3) are ways of saying the same thing in different varieties.

1. a. I’ve not finished it.
   b. I haven’t finished it.
2. a. I don’t know whom to trust.
   b. I don’t know who to trust.
3. a. I have a friend lives in Auckland.
   b. I have a friend who lives in Auckland.

In the United Kingdom, (1a) is more likely to be heard in the north (Scotland and northern England), (1b) in the south; (2a) is much more formal than (2b), with the result that (2b) is more likely to occur in spoken language and (2a) is often confined to some kinds of written language (Microsoft Word suggests changing (2b)). The sentence in (3a) illustrates a structure which can be heard, but is not generally accepted, and which is found more often in informal styles and in some dialect areas, while (3b) is the standard form (the form of the language which is deemed most prestigious and used in formal contexts). Different varieties may use different grammatical patterns (they may also be different varieties because they sound different – in Lancashire, England, wood may rhyme with cud – or use different words altogether – slater and woodlouse are both words to denote the same creature), so that a person from Perth, Australia, may use different grammatical patterns from a person from Perth, Scotland, a lawyer may use different grammatical patterns (speak a different variety) from a carpenter, a man may use different grammatical patterns from a woman, a woman may use different grammatical patterns in addressing her daughter and in addressing her grandmother, and so on.

One particular kind of variety is brought about by language change: things which used to be normal in English are now no longer normal. In the late 1800s, (4a) was the ordinary way of expressing the meaning, and (4b) did not exist, though today speakers would probably find (4a) very odd (see Strang 1970, pp. 98–9).

4. a. The house is building.
   b. The house is being built.

Language changes all the time, and once the time-depth is great enough, the old seems odd and the new seems normal. But while the change is actually occurring and both forms can be found, change can cause some confusion. Such periods of confusion can sometimes be very long, and then we often find that the speech of younger speakers is systematically different from that of older speakers. This confusion sometimes gives rise to statements of what ‘should’ be done, which can have the effect of prolonging the life of a moribund construction.

Descriptive and prescriptive grammar

Notions of what should be done in grammar are called ‘prescriptive’ or ‘normative’ statements. Prescriptive ideas, ideas about the way language ought to be, have several origins. In some cases, a grammarian will describe what he or she believes is the case in the language of people who are thought to use the ‘best’ English – often literary writers, biographers and historians – and this gets interpreted as meaning that everyone should copy this usage.

In some instances, the copied usage is a minority one. For example, it is probably the case that for the majority of English speakers around the world (though more often in some areas than in others), I never did it last week is a perfectly normal statement, even though, in standard English, never is incompatible with last week, and I didn’t do it last week would be the expected form.

In other cases, it may be suggested that English grammar should follow a pattern that is found in some other language (a language with high cultural prestige, often Greek or Latin). For example, the class of prepositions is so-called because in Latin, the corresponding set of words occurred before (pre-) noun phrases. In English, they are also found in that position, but the same forms also have another use in expressions like come out, drop in, put up, and in other places where they do not always occur before a noun phrase. Lowth (1762) knew that what we now call ‘preposition stranding’ was ‘an Idiom which our language is strongly inclined to’ [sic] (quoted in Huddleston & Pullum 2020, p. 202), and it was later authors who tried to impose what they perceived as a Latin pattern on English and left us with awkward normative statements such as ‘Never end a sentence with a preposition’ (to which, of course, the response may be ‘What would I want to end a sentence with a preposition for?’ or ‘What would I want to use a preposition to end a sentence up with for?’).

In a third set of cases, grammarians may describe an older pattern and readers may take it as implying (whether that is the author’s intention or not) that the newer pattern replacing it should not be used (as is illustrated with whom in (2a)).

By contrast, so-called ‘descriptive’ statements about language claim to do no more that state what grammatical patterns can be observed in usage. They may also try to provide explanations of the observed patterns, and the more modern grammars may indicate where there are differences of opinions on such matters. The three major grammars cited above are all descriptive in this sense, but all descriptive works may end up being used normatively, for example in second-language teaching situations.

What becomes of interest then, is how people know what general usage is, and how to identify appropriate data for description. Many writers make up their examples (as we have done to this point), and assume that their readers will agree with them about what is or is not part of English. In fundamental sentences like (5), that may well be true, but there are plenty of instances where we might disagree, and the disagreement may be minor or major.

5. The cat sat on the mat.

For example, in the third decade of the twenty-first century there is likely to be disagreement about the sets in (6)–(7); even if individual speakers are consistent in their own usage, there may be disagreement between speakers.

6. a. I’m bored with this game.
   b. I’m bored of this game.
7. a. Have you any money?
   b. Do you have any money?
   c. Have you got any money?

Sources of evidence

To avoid the bias of citing one’s own personal intuition – which, in some cases, may change from day to day – many grammarians prefer to cite examples of usage from users who have, as far as we can tell, not been consciously manipulating their usage when they produced the relevant material (and we do this quite often in this book). It is possible, of course, for people to make mistakes in their usage – or at least, to say things which they would wish to correct if they had more time. Making mistakes is particularly easy to do when speaking as opposed to writing, but it can happen in any language medium. To avoid such errors, we would ideally like to see a particular pattern repeated from several authors, but it is often difficult to find suitable examples. To maximize the chances of getting good data, linguists and grammarians these days frequently use corpora.

A corpus is a large body of data selected in such a way as to provide usage from many people, and entered into a computer database to allow for easy searching. Corpora can be anything from a few thousand words of running text to billions of words of running text, depending on when the texts in the corpus were collected and how they were gathered and treated. Typically, for example, corpora of spoken English are smaller than corpora of written English, because a lot more work is involved in entering and editing data for a corpus of spoken language than for one of written language. Older corpora (from the late twentieth century), which took material from printed media, tend to be smaller than modern corpora which take material from internet sources. We provide examples from a number of corpora in this book, and the corpora used are listed in our set of references, along with the books that we have consulted.

Some corpora are parsed, which means that each word is assigned to a word class, such as noun, verb, adjective and so on. Because most of the parsing is done automatically, using artificial intelligence and statistical techniques, there are to date still problems involved in doing this. Though it is rarely completely accurate, in principle it makes it easy to distinguish between standard (‘a flag’) and standard (‘normal’), or to see what nouns fit into the noun slot in a high [Noun] for (for example, we can have high hopes/regards/praise for, but not, high corpus for). While corpora allow us access to a great deal of good data, the data has been democratized, in the sense that the data in such collections is typically the written usage of everyday writers, with no priority given to writers who are considered models of the best usage, as was typically the case in earlier descriptions, including dictionaries.

What is grammar?

We have not yet considered the meaning of the word grammar, a word that has an unfortunately large number of meanings among linguists and grammarians (Crystal 2015, Bauer 2021).

In one usage, grammar is the science of the study of all language phenomena. This might involve phonetics (is the [s] sound in cuts in She cuts the grass every week systematically different from the [s] sound in cuts in She has several bad cuts on her hand ?), phonology (are the rules for stressing nouns different from those for stressing verbs in English?) and pragmatics (can you give an order by making a statement?) – incidentally, the answers to all of these questions is ‘yes’.

A more restrictive use of grammar omits these factors, and keeps the word ‘grammar’ for two things: syntax (the way in which words are ordered and how that relates to the meaning of larger units) and morphology (the way in which meaningful elements go together to make up words, so that we can have wasp-ish-ness but not *wasp-ish-ity or *wasp-ity-ish). (The asterisk is a device conventionally used by linguists to show that what follows is not good English.)

This latter is the main sense in which we will use the word. A grammar (not just grammar any more) will also be used, as it has already been used in the second example, to denote a description of a language or a part of a language focusing on these questions.

Grammars, in this sense, focused heavily on regularities in the language. Their interest was to uncover and document regular patterns in the language. A typical example of such a regular pattern might be how verbs are adjusted for different persons (first, second, third). The left-hand side in (8) gives the Latin version of the how the verb meaning ‘love’ is ‘conjugated’ and the right-hand side gives the English version.

8. a. amo‘ I love’
   b. amas‘ you love’ (just one person)
   c. amat‘ he/she/it loves’
   d. amāmus‘ we love’
   e. amātis‘ you love’ (more than one person)
   f. amant‘ they love’

You can immediately spot that Latin m...