Previous PageTable Of ContentsNext Page

Introduction to English Language Online

Return to IEL Home Page | Comments on Activities

Unit 9: Grammar and the Corpus

Activity 1

Each line of text in the set of lines below comes from a separate text. The gap down the middle represents a missing word, the same word in every line. Find the word, and make a careful note of precisely what clues led you to the answer. The clues may be grammatical, lexical or semantic.

<is all delivered rolled up and then

 

deploys as a result of various>

<national conferences to usher out,

 

without bloodshed the old one->

<human energetics in general.

 

the background information will>

<Co-operation in Europe (CSCE)with,

 

, Nato playing a steadily>

<years ago he often used words like

 

or phrases like `no way'. Many>

<mixed race guy, for friendship,

 

more. Photo if poss. A.L.A. Write>

<liaison process instituted to

 

prevent further similar mistakes.

<fee to a local hospital charity.

 

, this humanitarian act took the >

<21+ with car for friendship,

 

1-2-1 relationship disabled or>

<radio studio, huh?<t> Young: Well,

 

your listeners will continue to>

<years out of college, who had had,

 

, some management experience, who>

<Her wide, bald-looking eyes stared

 

out of her pale face, but still>

<will do The Spirit won't they. Yeah

 

they might do something else>

<MX and then we can get into a more

 

more kind of er erm raw blooded>

<want to prove everybody wrong, and

 

I can help keep Southampton in >

<He would be in his forties now and

 

with a son or two of his own to>

<said: `I've had to be patient, but

 

his can be the start of my Leeds

<gladly comply with its demands.

 

we would not all be blamed for >

<and the festival circuit moves on,

 

to cheaper places, with fewer>

<stressful on the mind and body but

 

everything will be all right," he>

<concerts from Madonna and U2 would

 

be confirmed in the near future,">

<before # It was a nightmare which

 

I'll never repeat # >

Activity 2

Word Class Revision.

Look at the lines below and try to divide the lines up according to which word class round belongs to. Does round belong to any other word classes not represented here?

< list featuring Italian selections round out a meal that can only be >

< then capture the probe, swinging it round and accelerating it South till it >

< convincingly in the second round of voting for the leadership of the

< the car beside him. Then she looked round in surprise. `Where's Denis, Sam? I

in bewilderment. Adeline put her arms round her to hide her face for a moment. >

< <FCH> Pan, 1971) but it has all round value. Its properties seem to >

< refreshment can take heart from our round-up of pubs offering a welcome at >

<a viewing. Spend a few hours driving round doing a reconnaissance. You will >

< the screw by producing a clear round on Ricochet,which lifted him to >

< listed under the county name in our Round Britain Gay Guide (which should >

< or ten times a day as she dashes round town making dreams come true. She >

< says the two sides are due to hold round-table talks tomorrow # Rebels now >

< Here are some examples. Round and round the house went the crowd shouting >

< after <M01> Mm.<F01> sort of going round the docks and depositing its goods >

< t believe me and send the inspector round.<M01> Mm.<M02> Cos he's been round >

<you know they have people who go out round the supermarkets. The supermarkets >

<0 against Shildon Town in the second round in 1990. <t> St Albans play their >

think something so awful could happen round here." <t> Mr Bailey, 50, added: ` >

Frenchman Valery Kayumba in the ninth round of his European light-welterweight >

< when the draw for the second round of the Pilkington Cup was made at >

<of the world when I started my first round," he remarked.<t> Yet the new >

It creeps along eating bacteria." The round heads contain spores, and Calentine

< the idea of a coalition.<p> A round table might set an early date for >

< players back on deck after the last round loss to Brisbane # Coach Brian >

first round and postponing the second round to 1998 and effectively to Peter >

Two kinds of data

The grammar analysis which we have presented during this course is, despite appearances at times, a relatively simple and highly idealised model. Much of it depends on a long tradition of grammatical analysis stretching back over two thousand years. Data for constructing the model has always fallen into two broad categories. First there has been what grammarians thought the language consisted of, using their own introspective intuition. Second, there have been examples of language, written or spoken, but usually written, collected and recorded as a database. It is easy to see that the first kind of data does not necessarily agree with the second.

In recent years, the development of computer technology has enabled linguists to store and retrieve far more data than ever before. This means that the second kind of data has received far more prominence. One of the most important issues in language description is whether this new tool available to linguists leads to a fundamental reappraisal of the way language works.

The prevailing view in what is loosely known as 'Chomskyan' linguistics is that it does not. This is because a Chomskyan model relies on the principle of generation of a multiplicity of sentences from a relatively small set of basic rules. It argues that it is unnecessary to consult millions of examples of language, since the basic rules are unaffected by this. Analogies are often drawn with games such as snooker. A visitor from Mars, charged with the task of recording the rules of snooker by just observing people playing, would need to watch just a few games. After that, no matter how many different games or shots were observed, they would not reveal anything more about the underlying rules, which are known by the players.

Likewise, it is argued, the basic rules of language are subconsciously 'known' to any normal native-speaker, and it is the task of the linguist to explicate them. The linguist does not need more than a reasonable quantity of normal linguistic output as evidence, and for most purposes what is in the linguist's own head will do. (An obvious exception would be the description of dialects etc. where informants are needed.) It is argued, moreover, that at a deeper level the same rules account for the structure of all languages, not just English. The things which are unique to specific languages, such as vocabulary and the 'superficial' grammatical structure, have to be acquired by speakers of that language, but describing these can only ever be a secondary activity. This is the position first advanced by Chomsky and reaffirmed by Pinker.

An alternative view is that language is really quite unlike snooker, and that looking at very large quantities of language does indeed tell us something quite new. The following passage gives a good flavour of this view. The author is discussing early work in the development of corpora:

However, the availability nowadays of much larger corpora makes it possible to evaluate conventional grammatical statements. Presumably the shorter original corpora did little more than confirm the generally agreed positions on English grammar. The new evidence suggests that grammatical generalizations do not rest on a rigid foundation, but are the accumulation of the patterns of hundreds of individual words and phrases. The language looks rather different when you look at a lot of it at once.

(Sinclair 1990:100 Corpus, Concordance, Collocation. OUP)

Note that the discrepancy between what is contained in a large corpus and what grammarians sometimes say about the language can be illustrated in two separate ways. First, it can show how out of touch some prescriptive judgements are with what people actually say. Complaints against things like "split infinitives" (e.g. to boldly go) can be put in perspective by looking at the massive number of these in a corpus of ordinary English. The other, more technical, aspect of the discrepancy is in the area of grammatical description. For example, early studies of the corpus showed that some very common verbs in English, such as take or make, are used with very high frequency in patterns where they do not really have any meaning. Expressions like take a bath, or make a difference, are very common, but relatively fixed. It isn't possible to treat bath as the object of take in a normal clause analysis. If it were, we should be able to say things like a bath was taken by John, which seems impossible. Rather take a bath operates more as one three-word verb.

On the other hand, looking at data in the corpus requires considerable care. Simply noting a frequent pattern of occurrence of words one after the other does not necessarily tell you everything about the grammar of those words. It is usually necessary to have some intuitive feel for how the language is structured in order to interpret the data.

Activity 3

Consider this short extract from the COBUILD grammar (p.191). It is explaining that there is a group of verbs which take an object and a 'to' - infinitive clause. These have been picked out by the compiler by looking for the pattern of VERB + NP + 'to' + VERB. There are two example sentences demonstrating this pattern. In these, him is the object of taught and students is the object of encourage.

"3.207 Other verbs are used with an object and a

'to'-infinitive clause.

His sister had taught him to sew.

I encourage students to do these exercises at home.

Here is a list of verbs which are used with an object

and a 'to'-infinitive:

advise defy instruct pay tell

allow enable intend permit train

ask encourage invite persuade trust

beg expect lead prefer urge

cause forbid leave press use

challenge force like programme want

choose get mean prompt warn

command help move recruit

compel induce oblige remind

dare inspire order teach

Note that some of the verbs in the above list are used for reporting orders, requests and advice."

Arguably, some of the verbs in this list do not belong there. There are two reasons why this might be so:

1. Even though they can be used in the pattern V + NP + 'to' infinitive, the NP in such cases is not the object of the V.

2. Even though the NP is in fact the object of the V, the 'to' infinitive does not relate to the first V in the same way that the example sentences do; it means something like 'in order to'.

Which of the following sentences would be examples of the pattern the grammar is trying to describe at 3.207, and which are examples of different patterns?

(Activity 3 continued)

a. He advised me to leave.

b. He helped me to get over it.

c. He caused them to rebel.

e. He left her to finish the job by herself.

f. He uses us to smuggle the stuff in.

g. He likes you to be on time.

h. He left her to pursue his career as a priest.

i. He didn't mean you to stay.

j. He challenged me to a duel.

k. He taught us to think clearly.

Previous PageTop Of PageTable Of ContentsNext Page

Return to IEL Home Page | Comments on Activities