2. Words, sentences, and syntax Everybody knows what a word is – and everybody knows what a sentence is. This is all we’ll take for granted for the time being. However, not everybody knows what syntax is: it is the study of how words are combined into sentences. Now, if producing a sentence was just a matter of sticking two or more words together at random, you would be able to make 5.264.979.687 (!!!) different English sentences out of the following 12 English words: [2]
[a, button, by, of, on, radio, someone, that, the, touch, turned, was]
- giving you ‘sentences’ like these: [3]
a. a the b. a button c. someone turned d. touch the by was e. that the radio was by someone f. that the radio turned on was g. someone turned on the radio by a touch of that button etc., etc., etc, etc. etc. etc., etc., etc., .. etc
Now, your intuition will tell you that not all of these are real English sentences. But some of them are: [c], [e] and [g]. Perhaps [g] is a little ‘better’ than [e] - but then [e] is far ‘better’ than [f], which also consists of six words. And [c] is perfectly OK, although it only consists of two words, like [a] and [b]. Of these two, [b] seems to be - well, English, but not an English sentence, only an English-sounding combination of sorts. In the same way, [f] is not necessarily excluded as a legitimate sequence of English words. In He said that the radio turned on was very expensive, [f] is part of a larger sequence, which is an English sentence. You may nevertheless ‘feel’ that somehow [b] is ‘better’ or ‘more complete’ than [f] - and you would be right. So obviously, there is more to it than just sticking words together in some random way. There would seem to be some kind of rule, or rules, that help you decide what combinations count as sentences ([c], [e], [g]), what combinations count as something ‘complete’, yet not a sentence ([b]), what combinations count as legitimate, yet ‘incomplete’ ([f]), and what combinations are gibberish and couldn’t possibly be part of an English sentence ([a], [d]). Syntax is concerned with finding, describing and explaining these rules. 2.1. Types of words: word classes Look at [3 a, b, c] again. All three consist of two words – [a] is gibberish, [b] is English, but not a sentence, while [c] is an English sentence. What does that tell us? It tells us that words are different. While two words like someone and turned may combine into a sentence, and two words like a and button may combine into something which is English but not a sentence, two words like a and the may not combine into anything. The reason must be that we recognize these words as being somehow different from each other. Part of the heritage from the classical tradition is the division of the words of a language into word classes. Some of them are well known (English terms in bold face, Classical terms in Roman type, and Danish terms in italics):
[4]
a. noun (N) b. verb (V) c. adjective (A)
(substantiv, navneord) ( verbum, udsagnsord) ( adjektiv, tillægsord)
These three word classes are also called the major lexical categories - ‘lexical’ clearly has something to do with ‘lexicon’, which is a list of words, and ‘category’ is another word for ‘class’. The major lexical categories are open; if the need arises, new nouns, verbs and adjectives may be ‘invented’ - by scientists, advertisers, novelists, poets, marketing people, and by quite ordinary people, especially children. The major lexical categories contain the words that we intuitively feel carry ‘real’ meaning - that is, meaning that somehow connects language with the real world: the meaning of the word ‘house’ is not in itself a house, of course. But because the word ‘house’ has the meaning it does have, we can use it to talk about houses. For this reason, the kind of meaning carried by members of the major lexical categories is often called descriptive. To say that some lexical categories are major suggests that there are in fact also minor lexical categories - and indeed there are. There is not the same firm tradition behind these, however. Those we shall use in this compendium are set out in [5] a. pronoun (Pro) b. determiner (Det) c. particle (Prt) - preposition (P) - conjunction (Conj) - adverb (Adv) d. auxiliary (Aux) e. proper name (PN)
(pronomen, stedord) (determinativ) (partikel) (præposition, forholdsord) (konjunktion, bindeord) (adverbium, biord) (hjælpeverbum) (proprium, egennnavn)
Apart from proper names and adverbs, all of the minor lexical classes are closed. In practice, this means that nobody can invent a new pronoun, conjunction, or preposition to meet a particular need. Notice the attempt, for example, to ‘invent’ a new pronoun, s/he, to replace he (or she) in order to eliminate ‘sexist’ language. It hasn’t really caught on, has it. And what form should replace him (or her)? Hir, or herm? The minor lexical categories contain words, just like the major categories. However, the meaning associated with members of the minor lexical categories is of a far more abstract kind than that associated with nouns, for example. The meaning of the determiners, the prepositions, and the conjunctions does not connect with anything in the real world in the way that the word house connects with real houses. The kind of meaning associated with the minor lexical categories plays an entirely different role. It helps both the speaker and the listener to structure the meaning carried by major lexical elements. For that reason, this kind of meaning is often also called structural. 2.1.1. Proper names are an exception to this. They are the names of people, teddy bears,
places, horses, dogs, the family car, star constellations, and any other individual that people for some reason want to give a name of its own. So the semantic characteristic of proper names is that they connect with particular things in the real word. We won’t have much to say about proper names. 10
2.1.2. Pronouns (some of them) share with proper names the semantic characteristic of
connecting with particular things in the real world. Words like I and you connect with the speaker and the listener, respectively, and sometimes he, she, it connect with the person or thing spoken about. But they differ from proper names in always being dependent on the actual utterance situation for their interpretation. The connections of other pronouns, like who, which, that, is a bit more tricky, but we won’t have much to say about them this term. 2.1.3. The class of adverbs is quite special, mainly because it is the traditional ‘ragbag’ of words that don’t fit clearly into any other category. There are two basic kinds of adverbs:
[6] a. ‘Born’ adverbs: words like here, now, then, sometimes, perhaps, not, whenever, too, etc. b. Derived adverbs: words like clearly, beautifully, sincerely, etc. that are formed (‘derived’) from an adjective by means of the ending -ly. Given that every adjective can be the ‘mother’ of a corresponding adverb by means of -ly, and given that the adjectives form an open category, it follows that the category of adverbs is also ‘open’. But this only works for the derived adverbs. The ‘born’ adverbs form a closed category. Apart from those in [6], there are words like in, on, out, down, off, etc., which sometimes behave like adverbs but which are usually classified as prepositions, words like because, when, yet, which sometimes behave like adverbs, but are usually classified as conjunctions, and finally words like before, since, which sometimes behave like adverbs, sometimes like prepositions, and sometimes like conjunctions. The fact that these are difficult to classify is the main reason for recognizing a ‘superclass’ of particles, comprising adverbs, prepositions and conjunctions.
11
Topic for Discussion: Classification The traditional word classes are the result of classification. Think of classification as the process of sorting a pile of something, eg. fruit, into smaller, uniform piles. Now, what should the result of this sorting be? That depends on its purpose. The crucial thing about classification is that its purpose will determine one or more relevant criteria by which the sorting should be carried out. In the case of fruit, if your purpose is to paint a nature morte, you might make colour the relevant criterion. If on the other hand you are a keeper in the monkey house of a ZOO, needing to feed your inmates, you might make weight the relevant criterion, and so on. And notice that only a combination of criteria (eg. colour, size, shape, taste, smell) will help you sort all the fruit into piles according to their kind. In the case of words, what relevant criteria might there be? Might it, for example, be relevant to find out how many letters they are made up of? It could be - if your purpose was to set up a list of English words that you should not use as a foreigner. These would be swear-words (also called four-letter words). But usually the number of letters is not considered relevant for a grammatical classification. Instead, the following three types of criteria might be: 1. Form: Are there any formal characteristics of English words that may serve as criteria for classification? Yes, there are. We have just seen that some adverbs are formed by adding -ly to an adjective. This is an example of a formal characteristic. However, as we also saw, not all adverbs end in -ly. Therefore this characteristic would be too narrow a criterion to help us set up the entire class of adverbs. Another formal characteristic might be the ability of some English words to take the ending -s, eg. talk - talks. But that would be too broad. We cannot tell if talk and talks are nouns or verbs without further information. We could add another two characteristics to this one, and say that every English word that can take the endings -s, -ing, and -ed belongs to the same category, the class of verbs. This would be true, but it would again be too narrow. It would be enough to help establish the class of regular verbs in English. Yet there are many words which we - for different reasons - would want to include in the class of verbs that would not be covered by this set of criteria, eg. be, have, do, can, break, steal, etc. etc. In general, formal criteria cannot be used alone to establish classes of English words. 2. Meaning:
Is there an element of meaning that all words of a given class have in common? Well, nouns are often said to connect with things, verbs with processes, and adjectives with qualities. However, while it is no doubt true that all words that do connect with things are nouns, the opposite isn’t necessarily true. What kind of ‘thing’ is chemotherapy, love, angel, and talk? In general, meaning cannot be used alone to establish classes of English words.
3. Function:
Compare I heard the talk and I heard them talk. If you heard these two sentences spoken, the only difference you would hear would be the presence vs. absence of /m/ just before /tɔ:k/. Nevertheless, this would be enough for you to interpret talk in two different ways: in the first example, talk is (what grammarians call) a noun, in the second it is a verb. It would seem that the ‘same’ word just has two different functions in these two sentences. Nevertheless, we do recognize two words, a noun and a verb. This is because the functional criterion is supplemented by formal ones. In general, the most important criterion for establishing the classes of English words is function, but it is usually supplemented by criteria of form.
The interrelations between form and function will be one of the recurrent topics throughout. It is one of the crucial topics of grammar.
12
2.2. The formal characteristics of the major lexical categories Although formal characteristics cannot serve as the only criterion for the establishment of word classes, they are nevertheless important. For whenever a class has been established (on functional criteria) it will turn out that most of its members - including any new ‘inventions’ - will have the same formal characteristics. Thus - apart from a very few irregular ones - all nouns in English will have two phonetically distinct forms, and four orthographically distinct forms (here illustrated with the noun dog): Phonetic [7] a. /'dƆg/ b. /'dƆgz/ c. d. -
Orthographic dog dogs dog’s dogs’
Main function singular plural genitive singular genitive plural
In the same way, most verbs will have four phonetically as well as orthographically distinct forms (here illustrated with the verb destroy): [8] a. /dɪs' trɔɪ/
destroy
b. /dɪs' trɔɪz/ c. /dɪs' trɔɪd/
destroys destroyed
d. /dɪs' trɔɪɪɳ/
destroying
infinitive imperative present tense, 1st & 2nd singular present tense, plural present tense, 3rd singular past tense past participle present participle gerund
The tense forms here are called the finite forms, the italic forms are called non-finite. The imperative is neither finite nor non-finite. Notice that, strictly speaking, it is not the forms as such which are (non-)finite, but the functions of these forms. Lastly, many adjectives will have three phonetically as well as orthographically distinct forms: [9] a. /kǝʊld/ b. /kǝʊldǝ/ c. /kǝʊldɪst/
cold colder coldest
positive comparative superlative
2.3. Morphology
The oppositions between {singular | plural} in nouns, {present | past} in verbs, and {positive | comparative | superlative} in adjectives are examples of inflectional morphological distinctions. Morphology is the study of how word forms are composed from smaller, grammatically significant elements called morphemes. Inflection (‘bøjning’) is that part of grammar which lists the various grammatical forms of the same word. Obviously, inflection is more important in the grammar of languages like French, Spanish and German than it is in English, which has relatively little inflectional morphology. Danish falls somewhere between them.
