AN-244
Phrasal Syntax seminar
Marosán Lajos

Parts of Speech
Tarr Dániel

1995
Parts of Speech

Parts of Speech are words classified according to their functions in sentences, for purposes of traditional grammatical analysis. According to traditional grammars eight parts of speech are usually identified: nouns, adjectives, adverbs, prepositions, conjunctions, pronouns, verbs, and interjections. Noun girl, man, dog, orange, truth ... Pronoun I, she, everyone, nothing, who ... Verb be, become, take, look, sing ... Adjective small, happy, young, wooden ... Adverb slowly, very, here, afterwards, nevertheless Preposition at, in, by, on, for, with, from, to ... Conjunction and, but, because, although, while ... Interjection ouch, oh, alas, grrr, psst ...

Most of the major language groups spoken today, notably the Indo-European languages and Semitic languages, use almost the identical categories; Chinese, however, has fewer parts of speech than English.[1]

The part of speech classification is the center of all traditional grammars. Traditional grammars generally provide short definitions for each part of speech, while many modern grammars, using the same categories, refer to them as “word-classes” or “form-classes”. To preface our discussion, we will do the same:

Nouns
A noun (Latin nomen, “name”) is usually defined as a word denoting a thing, place, person, quality, or action and functioning in a sentence as the subject or object of action expressed by a verb or as the object of a preposition. In modern English, proper nouns, which are always capitalized and denote individuals and personifications, are distinguished from common nouns. Nouns and verbs may sometimes take the same form, as in Polynesian languages. Verbal nouns, or gerunds, combine features of both parts of speech. They occur in the Semitic and Indo-European languages and in English most commonly with words ending in -ing.

Nouns may be inflected to indicate gender (masculine, feminine, and neuter), number, and case. In modern English, however, gender has been eliminated, and only two forms, singular and plural, indicate number (how many perform or receive an action). Some languages have three numbers: a singular form (indicating, for example, one book), a plural form (indicating three or more books), and a dual form (indicating, specifically, two books). English has three cases of nouns: nominative (subject), genitive (possessive), and objective (indicating the relationship between the noun and other words).

Adjectives
An adjective is a word that modifies, or qualifies, a noun or pronoun, in one of three forms of comparative degree: positive (strong, beautiful), comparative (stronger, more beautiful), or superlative (strongest, most beautiful). In many languages, the form of an adjective changes to correspond with the number and gender of the noun or pronoun it modifies.

Adverbs
An adverb is a word that modifies a verb (he walked slowly), an adjective (a very good book), or another adverb (he walked very slowly). Adverbs may indicate place or direction (where, whence), time (ever, immediately), degree (very, almost), manner (thus, and words ending in -ly, such as wisely), and belief or doubt (perhaps, no). Like adjectives, they too may be comparative (wisely, more wisely, most wisely).

Prepositions
Words that combine with a noun or pronoun to form a phrase are termed prepositions. In languages such as Latin or German, they change the form of the noun or pronoun to the objective case (as in the equivalent of the English phrase “give to me”), or to the possessive case (as in the phrase “the roof of the house”).

Conjunctions
Conjunctions are the words that connect sentences, clauses, phrases, or words, and sometimes paragraphs. Coordinate conjunctions (and, but, or, however, nevertheless, neither … nor) join independent clauses, or parts of a sentence; subordinate conjunctions introduce subordinate clauses (where, when, after, while, because, if, unless, since, whether).

Pronouns
A pronoun is an identifying word used instead of a noun and inflected in the same way nouns are. Personal pronouns, in English, are I, you, he/she/it, we, you (plural), and they. Demonstrative pronouns are thus, that, and such. Introducing questions, who and which are interrogative pronouns; when introducing clauses they are called relative pronouns. Indefinite pronouns are each, either, some, any, many, few, and all.

Verbs
Words that express some form of action are called verbs. Their inflection, known as conjugation, is simpler in English than in most other languages. Conjugation in general involves changes of form according to person and number (who and how many performed the action), tense (when the action was performed), voice (indicating whether the subject of the verb performed or received the action), and mood (indicating the frame of mind of the performer). In English grammar, verbs have three moods: the indicative, which expresses actuality; the subjunctive, which expresses contingency; and the imperative, which expresses command (I walk; I might walk; Walk!)
Certain words, derived from verbs but not functioning as such, are called verbals. In addition to verbal nouns, or gerunds, participles can serve as adjectives (the written word), and infinitives often serve as nouns (to err is human).

Interjections
Interjections are exclamations such as oh, alas, ugh, or well (often printed with an exclamation point). Used for emphasis or to express an emotional reaction, they do not truly function as grammatical elements of a sentence.[2]

It is useful to make a distinction and consider words as falling into two broad categories; closed class words and open class words. The former consists of classes that are finite (and often small) with membership that is relatively stable and unchanging in the language. These words play a major part in English grammar, often corresponding to inflections in some other languages, and they are sometimes referred to as ‘grammatical words’, ‘function words’, or ‘structure words’. These terms also stress their function in the grammatical sense, as structural markers, thus a determiner typically signals the beginning of a noun phrase, a preposition the beginning of a prepositional phrase, a conjunction the beginning of a clause. Closed classes are: pronoun /she, they/, determiner /the, a/, primary verb /be/, modal verb /can, might/, preposition /in, of/, and conjunction /and, or/. Open classes are: noun /room, hospital/, adjective /happy, new/, full verb /grow, search/, and adverb /really, steadily/. To these two lesser categories may be added: numerals /one, first/, and interjections /oh, aha/; and finally a small number of words of unique function which do not easily fit into any of these classes /eg.: the negative particle not and the infinite marker to/.

Quirk and Greenbaum[3]point out the ambiguity of the term word, for words are enrolled in their classes in their ‘dictionary form’, and not as they might appear in sentences when they function as constituents of phrases. Since words in their various grammatical forms appear in sentences that are normal usage, it is more correct if we refer to them as lexical items. Thus, a lexical item is a word as it occurs in a dictionary, where work, works, working, worked will all be counted as different grammatical forms of the word work. This distinction however is not always necessary, for it is only important with certain parts of speech that have inflections; that is endings or modifications that change the word-form into another. These are nouns /answer, answers/, verbs /give, given/, pronouns /they, theirs/, adjectives /large, largest/, and a few adverbs /soon, sooner/ and determiners /few, fewer/.
A word may belong to more than one class; for example round is also a preposition /”drive round the corner”/ and an adjective /”she has a round face”/. In such cases we can say that the same morphological form is a realization of more than one lexical item. A morphological form may be simple, consisting of a stem only /eg.: play/, or complex, consisting of more than one morpheme /eg.: playful/. The morphological form of a word is therefore defined as composition of stems and affixes.

We assign words to their various classes according to their properties in entering phrasal or clausal structure. For example, determiners link up with nouns to form noun phrases /eg.: a soldier/; and pronouns can replace noun phrases /eg.: him/. This is not to deny the general validity of traditional definitions based on meaning. In fact it is impossible to separate grammatical form from semantic factors. The distinction between generic /the tiger lives/ and specific /these tigers/, unmarked and marked forms prove that.
Another possible assignment is according to morphological characteristics, notably the occurrence of derivational suffixes, which marks a word as a member of a particular class. For example, the suffix -ness, marks an item as a noun /kindness/, while the suffix -less marks an item as an adjective /helpless/. Such indicators help to identify word classes without semantic factors.

For the sake of completeness, it should also be added that a word also has a phonological and an orthographic form. Words which share the same phonological or orthographic “shape”, but are morphologically unrelated are called homonyms /eg.: rose (noun) and rose (past tense verb)/. Words with the same pronunciation are specified as homophones, and words with the same spelling are determined as homographs. Words which partake the same morphological form are called homomorphs /eg.: meeting (noun) and meeting (verb)/. There is also a correspondence between words with different morphological form, but same meaning. These are called synonyms. Of the three major kinds of equivalence, homonymy is phonological and/or graphic, and synonymy is semantic.

We have to go back to the distinction of closed-class items and open-class items, because this introduces a peculiarity of great importance. That is, closed-class items are ‘closed’ in the sense that they cannot normally be extended by the creation of additional members. For example, it is very unlikely for a new pronoun to develop. It is also very easy to list all the members in a closed class. These items are said to be constitute a system in being mutually exclusive: the decision to use one item in a given structure excludes the possibility of using any other /the book or a book, but not *a the book/. These items are also reciprocally defining in meaning: it is less easy to state the meaning of any individual item than to define it in relation to the rest of the system.

By contrast open class items belong to a class in that they have the same grammatical properties and structural possibilities as other members of the class (that is, as other nouns or verbs or adjectives or adverbs), but the class is ‘open’ in the sense that it is indefinitely extendible. New items can be created and no inventory can be made that would be complete. This ultimately affects the way in which we attempt to define any item in an open class; because while it is possible to relate the meaning of a noun to another with which it has semantic affinity /eg.: house = chamber/, one could not define it as not house, which is possible with closed class items /this = not that/.

However, the distinction between ‘open’ and ‘closed’ parts of speech or word classes must not be treated incautiously. On the on hand, it is not very easy to create new words, and on the other, we must not overstate the extent to which we speak of ‘closedness’, for new prepositions like by way of [4]are no means impossible. Although parts of speech have deceptively specific labels, words tend to be rather heterogeneous. The adverb and the verb are especially mixed classes, each having small and fairly well defined groups of closed-system items alongside the indefinitely large open-class items. So far as the verb is concerned, the closed-system subgroup is known be the well established term “auxiliary”...

Some mention must be finally made of two additional classes, numerals and interjections, which are common in the difficulty of classifying them as either closed or open classes. Numerals whether the cardinal numerals /one, two, three/, or the ordinal numerals /first, second, third/, must be placed somewhere between open-class and closed-class items: they resemble the former in that they make up a class of infinite membership; but they resemble the latter in that the semantic relations among them are mutually exclusive and mutually defining. Interjections might be considered a closed class on the grounds that those that are fully institutionalized are few in number. But unlike the closed classes, they are grammatically peripheral - they do not enter into constructions with other word classes, and they are only loosely connected to sentences with which they may be orthographically or phonologically associated.

A further and related contrast between words, is the distinction between stative and dynamic. Broadly speaking, nouns can be characterized naturally as ‘stative’ in that they refer to entities that are regarded as stable, whether these are concrete /house, table/ or abstract /hope, length/. On the other hand verbs and adverbs can equally naturally be characterized as ‘dynamic’: verbs are fitted to indicate action, activity and temporary or changing conditions; and adverbs in so far as they add a particular condition of time, place, manner, etc. to the dynamic implication of the verb.

But it is not uncommon to find verbs which may be used either dynamically or statively. If we say that “some specific tigers are living in a cramped cage”, we imply that this is a temporary condition and the verb phrase is dynamic in its use. On the other hand, when we say that “a species of animal known as tiger lives in China”, the generic statement entails that this is not a temporary circumstance and the verb phrase is stative. Moreover some verbs cannot normally be used with the progressive aspect /*He is knowing English/ and therefore belong to the stative rather than the dynamic category. In contrast to verbs, most nouns and adjectives are stative in that they denote a phenomena or quality that is regarded for linguistic purposes as stable and indeed for all practical purposes permanent /Jack is an engineer - Jack is very tall/. Also adjectives can resemble verbs in referring to transitionary conditions of behavior or activity. /He is being a nuisance - He is being naughty/.

The names of the parts of speech are traditional, however, and neither in themselves nor in relation to each other do these names give a safe guide to their meaning, which instead is best understood in terms of their grammatical properties. One fundamental relation is that grammar provides the means of referring back to an expression without repeating it. This is achieved by means of pro-forms. Participles and pronouns can serve as replacements for a noun /the big room and the small one/, more usually, however, pronouns replace noun phrases rather than nouns /their beautiful new car was badly damaged when it was struck be a falling tree/.

The relationship which often obtains between a pronoun and its antecedent is not one which can be explained by the simple act of replacement. In some constructions we have repetition, which are by no means equivalent in meaning /Many students did better than many students expected/. In some constructions repetition can be avoided by ellipsis /They hoped they would play a Mozart quartet and they will/. Therefore the general term pro-form is best applied to words and word sequences which are essentially devices for rephrasing or anticipating the content of a neighboring expression, often with the effect of grammatical complexity.

Such devices are not limited to pronouns and participles: the word such can described as a pro-form as there are pro-forms also for place, time and other adverbials under certain circumstances /M.is in London and J. is there too/. In older English and still sometimes in very formal English we find thus and so used as pro-forms for adverbials /He often behaved silly, but he did not always behave thus/so/. But so has a more important function in modern usage, namely to substitute with the ‘pro-verb’ do for a main verb and whatever follows it in the clause /He wished they would take him seriously for his ideas, but unfortunately they didn’t do so/. Do can also act as pro-form on its own /I told him about it - I did too/.

Some pro-forms can refer forward to what not been stated rather than back to what has been stated. These are the wh-items. Indeed, wh-words, including what, which, who and when, may be regarded as a special set of pro-forms /Where is M.? - M. is in London. - J. is there too/. The paraphrase for wh-words is broad enough to explain also their use in subordinate clauses /I wonder what M. thinks/. Through the use of wh-words we can ask for the identification of subject, object, complement or adverbial of a sentence /They [S] make [V] him [O] the chairman [C] every year [A]. - Who [S] makes him...?/.

Now that we have outlined the various aspects of Parts of Speech, according to traditional grammars, we will look at some other approaches and other specifications, without the sake of complexity, only to widen our views a little more on the subject:

Otto Jespersen[5]starts out from the point that all clauses consist of several words. One word is defined or modified by another word, which in turn may be defined or modified by a third word. This leads to the establishment of different ranks of words according to their mutual relations as defined or defining. In the combination “extremely hot weather” weather may be called a primary word or principal; hot is a secondary word or adjunct; and extremely is a tertiary word or subjunct. Primary and secondary words are superior in relation to tertiary words; secondary and tertiary words are inferior in relation to primary words. It is therefore possible to have two or more (coordinate) adjuncts to the same principal /that nice [A] young [A] lady [P]/.

The logical basis of this system of subordination is the greater or lesser degree of specialization. Primary words are more special (apply to a smaller number of individuals) than secondary words, and these in their turn are less general than tertiary words. The word defined by another word, is in itself always more special than the word defining it, though the latter serves to render the former more special than it is in itself. Thus in the sentence “very clever student”, student is the most special idea, whereas clever can be applied to many more men, and very, which indicates only a high degree, can be applied any idea. Student is more special than clever, though clever student is more special than student; clever is more special than very, though very clever is more special than clever.

It is a natural consequence of these definitions that proper nouns can only be used as principals, and while there are thus some words that can only stand as principals as expressing highly specialized ideas, there are other words that may be either primary and secondary words in different combinations /conservative Liberals - liberal Conservatives/. Further there are words of such general signification that they can never be used as primary words, like the articles.

His further definitions of parts of speech fall under the categories of substantives (=principals), adjectives (=adjuncts), adverbs (=subjuncts), verbs (=verbs not subject to conjugation), verbids (=participles and infinitives), predicatives (=‘mediate adjuncts’; {most commonly}a verb connecting two ideas in such a way that the second becomes a kind of adjunct to the first (the object)./Eg.: the rose is red/), objects (=primary words, but more special as well as more general than the first principal /eg.: an owl sees a bird/.), and pronouns (= a separate “parts of speech”, understood differently according to the situation in which they are used).

Lyons[6] starts out from distinguishing formal and nominal definitions. Nominal definitions of the parts of speech may be used to determine the names, though not the membership, of the major syntactic classes of English. Creating syntactic classes on ‘formal’ distributional grounds, with all the members of each of them listed in the lexicon, associated with the grammar, will mean that though not all members of class X will denote persons, places and things; most of the lexical items which refer to persons, places and things will fall within it; and if this is so we may call X the class of nouns. In other words we have ‘formal’ class X and ‘notional’ class A; they are not co-extensive, but if A is wholly or mainly included in X, then X may be given the label suggested by the ‘notional’ definition of A.

He also points out the necessity of considering the distinction between deep and surface structure and define parts of speech not as classes of words in surface structure, but as deep-structure constituents of sentences. The distinction between deep and surface structures is not made explicitly in traditional grammar, but it is implied by the assumption that all clauses and phrases are derived from simple, modally ‘unmarked’ sentences. It is asserted that every simple sentence is made up of two parts: a subject and a predicate. The subject is necessarily a noun (or a pronoun standing for a noun). The predicate falls into one of three types according to the part of speech which occurs in it: 1. intransitive verb, 2. transitive verb with its object, 3. the ‘verb to be’ with its complement. The object, like the subject, must be a noun, while the complement must either be an adjective, or a noun.

Deriving from these associations with particular parts of speech it is possible to determine traditional parts of speech or their function solely on the basis of constituent-structure relations. The class of nouns is the one constituent class which all sentences have in common at the highest level of constituent structure. The class of intransitive verbs is the only class which combines directly with nouns to form sentences. The class of transitive verbs combine with nouns and with no other class to form predicates. Be is the copula-class, since it combines with both nouns and the class of adjectives. This argument rests of course on the specific assumptions incorporated in the syntactic function of the parts of speech; namely the status of the copula or ‘verb to be’, and the universality of the distinction between verbs and adjectives.

‘To be’ is not itself a constituent of deep structure, but a semantically-empty “dummy verb” generated for the specification of certain distinctions (usually carried by the verb) when there is no other verbal element to carry these distinctions. Sentences that are temporally, modally and aspectually unmarked do not need the dummy carrier /M. is clever/. As for the distinction between verbs and adjectives it is traditionally referred to as to do with the surface phenomenon of inflection. The adjective, when it occurs in predicative position, does not take the verbal suffixes associated with distinctions of tense, mood and aspect, but instead a dummy verb is generated by the grammar to carry the necessary inflexional suffixes /M. is clever - *M. is clever-s/. The verb is less freely transformed to the position of modifier in the noun-phrase; but when it occurs in this syntactic position, unlike the adjective, it bears the suffix -ing /the clever man - the singing man - *the clever-ing man/. A distinction between stative verbs and verbs of action is also relevant to English. Stative verbs do not normally occur in the progressive form, while the majority of English verbs, which occur freely in the progressive are called verbs of action. This aspectual difference is matched by a similar difference in English adjectives. Most adjectives are stative, in the sense that they do not normally take progressive aspect when they occur in predicative position /M. is clever - *M. is being clever/, but there are a number of adjectives which occur freely with the progressive in the appropriate circumstances /M. is being silly now/. In other words, to be stative is normal for the class of adjectives, but abnormal for the verbs; to be non-stative is normal for verbs, but abnormal for adjectives. It is, however, the aspectual contrast which correlates with the notional definition of the verb and the adjective in terms of “action” and “quality”.

To follow this argument Huddleston[7]points out that nothing said about inflection requires that all the forms of a lexeme should belong to the same part of speech. The main problem area concerns the traditional non-finite forms of verb lexemes, participles, the gerund and the infinitive. A participle is said to be a “verbal adjective”, while the gerund and the infinitive are “verbal nouns”. According to traditional doctrine, a gerund like writing in She likes writing letters is a noun because it is the object of the verb like. This would lead traditional grammarians to classify together as nouns words which are syntactically very different. /Eg.: Writing the letters took some time - The writing of the letters took some time/. Instead of saying that both are nouns because they are subject of took, he suggests that we call writing a verb in the former case, because it is the head of the extended verb phrase, and call it a noun in the latter case, because it is a head of the noun phrase. Since the relation between this later type of noun writing and the stem write is lexical rather than inflectional, he calls this a “deverbal noun”, for it is derived by a lexical-morphological process from a verb stem.

The second problem area concerns possessives. In the traditional treatment of forms like John’s in John’s book is regarded as an inflectional form of the noun John but is also said to have the force of an adjective. This is easily resolved in the light of the analyses of ‘s as a clitic rather than an inflexional suffix: John’s is not syntactically a single word, not a form of John, so that the issue of whether a lexeme and a member of its paradigm belong to the same parts of speech does not arise.

References

Huddleston, R. - Introduction to the grammar of English . [ Cambridge University Press, 1984 ].

Jespersen, O. - Modern English Grammar on Historical Principles . (Vol. II.) [ Allen and Unwin, 1954, London ].

Lyons - Introduction to Theoretical Linguistics . [ Cambridge University Press, 1968 ].

Mc Cawley, J.D. - The Syntactic Phenomena of English . [ The University of Chicago Press, 1988 ].

Quirk & Greenbaum - A Comprehensive Grammar of the English Language . [Longman, 1983, London ].

Quirk & Greenbaum - A University Grammar of English . [ Longman, 1973 ].

Quirk & Greenbaum - A Student’s Grammar of the English Language . [ Longman, 1991 ].

Microsoft (R) Encarta. 1994 Microsoft Corporation. [ Funk & Wagnall's Corp., 1994 ].
-----------------------
[1] See: Huddleston, R. - Introduction to the grammar of English . pp.90. [Cambridge University Press, 1984].
[2] See: "Parts of Speech," Microsoft (R) Encarta. 1994 Microsoft Corporation. [Funk & Wagnall's Corp.1994].
[3] See: Quirk & Greenbaum - A Comprehensive Grammar of the English Language . pp.68. [Longman, 1983].
[4] See: Quirk & Greenbaum - A University Grammar of English . pp.20. [Longman, 1973].
[5] See: Jespersen, O. - “The Three Ranks - Parts of Speech - Word Groups” in Modern English Grammar on Historical Principles . pp.65-74. [Allen and Ullwin, 1954]
[6] See: Lyons - “Parts of Speech” in Introduction of Theoretical Linguistics . pp.317-333.
[7] See: Huddleston - “The Parts of Speech” in Introduction to the Grammar of English . pp.90-122.