Like most language myths this one begs a number of questions, such as the following:
What is meant by ‘bad grammar’?
What is meant by ‘grammar’?
Can particular sentences of the English language reasonably be described as ‘slovenly’ – or ‘lacking in care and precision’, according to one dictionary definition?
The quest for answers exposes the myth to critical scrutiny.
Newspaper features, letter columns and the mailboxes of the BBC are good places to find complaints about bad grammar. A rich harvest may be gathered if language use becomes the subject of public debate or if current educational policies are focusing on English teaching and testing. In Britain recently many judgemental remarks have been aired about ‘Estuary English’, the name given to a variety of the language which is spreading both socially and geographically.
Examples of specific constructions often described as bad grammar can be placed in at least three categories. The first, exemplified in sentences (1)–(3) along with the (presumed) correct form in italics, regularly occur in the speech and writing of educated people.
(1) Who am I speaking to? / To whom am I speaking?
(2) Martha’s two children are completely different to each other. / Martha’s two children are completely different from each other.
(3) I want to quickly visit the library. /I want to visit the library quickly.
Two well-known ‘errors’ appear in (1), namely the preposition in the sentence final position and the nominative form of the relative pronoun ‘who’ rather than the oblique form ‘whom’ which is prescribed after a preposition. In (2) the expression ‘different to’ is used rather than the prescribed ‘different from’; and in (3) there is a ‘split infinitive’. In fact, the ‘correct’ versions were prescribed as such relatively recently in the history of the language, as part of the flurry of scholarly activity associated with the codification of the English language in the eighteenth century. Since the goal of codification is to define a particular form as standard, this process entailed intolerance of the range of choices which speakers and writers had hitherto taken for granted. In earlier centuries all these ‘errors’ appeared in highly sophisticated writing; in 1603, for example, Thomas Decker wrote ‘How much different art thou to this cursed spirit here?’
Different rationalizations were introduced to support these new prescriptions. The model of Latin was invoked to argue that a preposition should not end a sentence, that the inflected form of who should not appear anywhere other than in the subject of the sentence, and that an infinitive should not be split. The reason advanced by one writer of a popular manual of correctness for preferring ‘different from’ is that ‘different to’ is illogical, since no one would say ‘similar from’. But it is not difficult to construct an equally logical argument in support of ‘different to’, since it falls into a set of words with comparative meanings such as similar, equal, superior, which require to. Not only are prescriptive arguments difficult to sustain, but if taken seriously they are likely to create problems. For example, ‘Who am I speaking to?’ is normal in most contexts, while ‘To whom am I speaking?’ will generally be interpreted as marking social distance. Thus the real difference between these forms is stylistic; both are good English sentences in appropriate contexts. Sometimes an attempt to follow the prescribed rules produces odd results.
(4) A good author needs to develop a clear sense of who she is writing for.
(5) A good author needs to develop a clear sense of for whom she is writing.
The prescription, which outlaws (4) and yields (5), does not work because it is not based on a principled analysis of the structure of English but is a response to cultural and political pressures. By the eighteenth century Britain needed a standardized language to meet the needs of geographically scattered colonial government servants and to facilitate mass education. It did not too much matter which of a set of variants emerged as standard, as long as only one was specified as such. The prescribed standard was codified in grammars (such as Robert Lowth’s) and dictionaries (the most famous being Dr Johnson’s). No systematic grammar of English existed at that time, but Latin had a particular prestige as the lingua franca of scholars throughout Europe; hence the appeal not only to logic but to the model of Latin to justify particular prescriptions. But as we shall see shortly, English rules are very different from Latin rules, though equally complex; like all Germanic languages, English quite naturally places prepositions in sentence final position.
By ‘bad grammar’ then is sometimes meant expressions which are not in line with even unrealistic prescriptions. But what is grammar? Our myth refers to a prescriptive grammar, which is not a systematic description of a language, but a sort of linguistic etiquette, essentially an arbitrary set of dos and don’ts. Two other kinds of grammar can be distinguished – a descriptive grammar and a mental grammar.
A descriptive grammar does not set out to legislate on correctness but describes how words are patterned to form major constituents of sentences. The distinctive rules of English which underlie these patterns are acquired by children and learnt by speakers of other languages but are generally taken for granted by prescriptive grammars. One basic rule of this type describes how questions are formed in English. Consider the following declarative sentences and their corresponding questions:
(7) Martha is cooking lasagne for dinner tonight.
(8) Martha should have cooked lasagne for dinner tonight.
(9) My new flatmate who has won the Cordon Bleu cooking contest is celebrating with a party tonight.
(10) Martha cooks lasagne every Friday.
(6a) Is Martha Peter’s sister?
(7a) Is Martha cooking lasagne for dinner tonight?
(8a) Should Martha have cooked lasagne for dinner tonight?
(9a) Is my new flatmate who has won the Cordon Bleu cooking contest celebrating with a party tonight?
(10a) Does Martha cook lasagne every Friday?
Sentence (6) makes the rule seem simple: the verb moves to the beginning of the sentence. However, (7) and (8) contain a complex verb phrase, consisting of a lexical verb and one or more auxiliary verbs. Lexical verbs can be identified as dictionary entries and form an almost infinitely large class which is constantly being augmented by new borrowings and inventions: examples are kick, sneeze, vegetate, computerize, shuttle, chortle, debug (some of these words can behave both as verbs and nouns). The set of auxiliary verbs, however, is sharply limited: it consists of forms of he (such as is and are)’, forms of have-, forms of do, and modals such as must, should, might, could, can, will. As (7) and (8) show, the question-formation rule moves the first auxiliary verb to the beginning of the sentence. But (9) shows that this does not always work. Simply moving the first auxiliary verb produces a truly ungrammatical sentence which is not in line with the rules used by any speaker of any kind of English. This kind of sentence is conventionally marked with a star:
(11) *Has my new flatmate who won the Cordon Bleu cooking contest is celebrating with a party tonight?
To solve this problem we need to modify our rule; the subject of the sentence changes places with the next auxiliary verb, which ends up at the front of the sentence. The subject of (9) is shown in italics:
(12) My flatmate who has won the Cordon Bleu cooking contest is celebrating with a party tonight.
Subjects of sentences can be of different kinds; they can be single words (she, Martha, flatmates); different sizes of noun phrases (my [new] flatmates [from Italy]). The subject of (9) is rather large and cumbersome, but we can formulate the question rule coherently only if we recognize that the subject of (9) is the whole sequence my new flatmate who has won the Cordon Bleu cooking contest and not some part of it. But complicated as this rule has become, it still needs some fine-tuning. Otherwise, how do we handle a sentence like (10), which does not contain an auxiliary verb? For some centuries English speakers have not formed questions in the manner of Othello (‘Thinkst thou I’d make a life of jealousy…’) by moving the lexical verb to the front of the sentence. In such cases as (10) we need to supply the appropriate form of do, as shown by (10a). But this also leads us to a yet more complex specification of our rule. Consider the ungrammatical sentence (13):
(13) *Do Martha cooks lasagne every Friday?
The verb cook in (10) is inflected with -s to mark present tense, and that present-tense marker must attach itself only to the auxiliary verb do in order to form the fully grammatical question (10a).
All the operations described above are required in order to construct a grammatical question. We here use the term ‘grammatical’ in a sense very different from that suggested by the prescriptive expression ‘bad grammar’. A grammatical sentence in this more technical sense follows the rules of the language as it is used by its native speakers. These rules are followed unconsciously and, generally speaking, native speakers do not make mistakes of the kind illustrated by (11) and (13). However, young children take some time to acquire rules; one three-year-old asked, ‘Did baby cried last night?’ Second-language learners and speakers with strokes or head injuries can certainly experience problems with the grammar of questions. Although this rule is very much more complicated than the list of dos and donts which are the focus of prescriptive grammars because knowledge of it is unconscious, many speakers who are familiar with the rather different rules of prescriptive grammar simply do not know that it exists. This unconscious knowledge of a set of rules (we have looked at only one of these) which allows native speakers to produce grammatical sentences and to distinguish grammatical from ungrammatical sentences (we did this when we considered (11)) can be described as a mental grammar.
Prescriptive rules are never as complex as properly formulated descriptive rules, and are easily dealt with by descriptive grammars. For example, different from/to would simply be specified as options; the split infinitive would not be an issue since the infinitive form of the verb is visit, not to visit; ‘Who am I speaking to?’ would be viewed as a normal sentence following the rules of English.
Sentences like (14) and (15) are also subject to popular criticism:
(14) So I said to our Trish and our Sandra, ‘Yous wash the dishes.’
(15) Was you watching the game when the rain started?
Unlike (1)–(6), which are regularly used by educated speakers and writers, both of these are characteristic of low-status speakers. They were recorded respectively in Belfast and London, although the grammatical patterns which they illustrate are found elsewhere. It is the low social status of these speakers, indexed by details of their language use, which seems in this case to form the basis of negative evaluation. In such a way is social class prejudice disguised as neutral intellectual commentary, and for this reason one linguist has described linguistic prescriptivism as the last open door to discrimination. But note that (14) makes a systematic distinction between ‘you’ (singular) and ‘yous’ (plural) similar to many languages of the world but lacking in Standard English. Thus, (14) cannot be argued to be in any sense linguistically impoverished (another common rationalization in defence of prescribed variants). Languages and dialects simply vary in the meaning distinctions they encode, regardless of their social status.
Note that (15) is a perfectly formulated question. Earlier in the history of English was and were in such sentences were acceptable alternatives (recall that the process of standardization has narrowed the range of socially and stylistically acceptable linguistic choices). But if we ask whether such sentences are ‘slovenly’ (‘lacking in care and precision’) we must surely concede that the care and precision needed to implement the question-formation rule is considerable, placing in perspective the triviality of requiring were with a plural subject.
Let us look finally at two sentences which seem to be subject to criticism for yet a different reason:
(16) Me and Andy went out to the park.
(17) it’s very awkward/it’s difficult mind you/with a class of thirty odd/occasionally with the second form/you’ll get you know/ well we’ll we’ll have erm a debate/
Neither (16) nor (17) are clearly marked as belonging to a particular region, but between them they display a number of characteristics of informal spoken English. Uttered by an adolescent boy, (16) is criticized on the grounds that the wrong pronoun case (me instead of I) is used inside a conjoined phrase. Speakers are so conscious of this Latin-based prescription that even linguistically self-conscious and quite prescriptively minded individuals sometimes hypercorrect and use I where me is prescribed (a particularly large number of complaints about these patterns of pronoun use are received by the BBC). Thus Margaret Thatcher once announced, ‘It is not for you and I to condemn the Malawi economy,’ and Bill Clinton pleaded, ‘Give Al Gore and I a chance.’ But a systematic analysis of English grammar reveals underlying rules which permit variation between me and I only within conjoined phrases. Thus, adolescent boys do not habitually say ‘Me went out to the park,’ Clinton would not plead ‘Give I a chance,’ and not even Margaret Thatcher would have said ‘It is not for I to condemn the Malawi economy.’ With respect to prescriptive rules, there is often such a disparity between what speakers believe is correct and what they actually do; but descriptive rules are neither subject to violation nor are they part of our conscious knowledge of language.
Although conversation is often thought to be unstructured, ungrammatical and slovenly (presumably when judged against the norms of writing or formal speech), its complex organizational principles are quite different from those of planned spoken or written discourse; it is not simply spoken prose. Transcribed from a coffee-break conversation between two teachers, (17) is typical of informal conversation in its chunks (marked by slashes), which do not correspond to the sentences of written English. Also in evidence are fillers such as erm, hesitations (marked by full stops), repairs, repetitions, and discourse tokens such as you know, mind you. Most of these features are attributable to conversation’s interactive, online mode of production, and the two discourse tokens function as ‘participation markers’, signalling to the interlocutor that interactional involvement or response is expected. Thus, it hardly seems appropriate to describe even the apparently unstructured utterance (17) as ‘slovenly’.
So what are we to say in conclusion about our current myth? ‘Bad grammar’ is a cover term to describe a number of different kinds of English expressions. Some are widely used by educated speakers and writers but are outlawed by traditional prescriptions which are difficult to sustain; some appear to attract covert social prejudice by virtue of their association with low-status groups; and some follow the very characteristic but still rule-governed patterns of informal speech. All are perfectly grammatical, providing evidence of a complex body of rules which constitute mental grammars, the unconscious knowledge which speakers have of their own language. In comparison, the prescriptions which are recommended as ‘good grammar’ are revealed as at best marginal and frequently as unrealistic and trivial.
For details of the processes and consequences of prescriptivism, see Rosina Lippi-Green, English with an Accent (London: Routledge, 1997) and James Milroy and Lesley Milroy, Authority in Language (London: Routledge, 1985). For a humorous critique of some common prescriptions see Patricia O’Conner, Woe is I: the grammarphobes guide to better English in plain English (New York: Grosset/Putnam, 1996). Mental grammars are discussed by Steven Pinker, The Language Instinct (London: Penguin, 1994) and for a standard descriptive grammar of English see Sidney Greenbaum, The Oxford English Grammar (Oxford: Oxford University Press, 1996).