1. General problems involving parts of speech
2. What is Construction Grammar?
2.1 The concept of construction outside the Construction Grammar framework
2.2 The concept of construction within Construction Grammars
2.3 On Construction Grammar
3. Construction Grammar and parts of speech
3.1 The general approach of Construction Grammar towards parts of speech
3.2 The Radical Construction Grammar approach towards parts of speech
This term paper is concerning Construction Grammar and the way it attempts to handle parts of speech categories like nouns, verbs and adjectives. The seminar this paper was written for proved in a manifold of ways that when we restrict ourselves to the categories provided by traditional grammar, we may face serious difficulties which call into question the fundamental categorizations of such grammar. These problems, or potential problems, are the subject matter of the first part of this paper.
In the second part, I will give an overview of Construction Grammar. Since the concept of construction is central to Construction Grammar, I will first clarify the notion of construction outside of the Construction Grammar framework in 2.1., before moving to an account in 2.2 of those features that can be seen as the smallest common denominator for defining constructions within different Construction Grammars.
One of my principal findings is the differing manners in which Construction Grammars in general, and a specific variant of Construction Grammar, namely Croft’s Radical Construction Grammar, address the issue of parts of speech. Part three will present Croft’s grammar as an answer to one of the central questions raised in this seminar, namely, which model is most adequate for categorising word classes in a single language like English but also in a cross-linguistic sense. This examination will be followed by some concluding remarks in part four.
1. General problems involving parts of speech
“Word classes (also know as parts of speech) are essential for any grammatical description, even though we can never really be entirely sure what their nature is. The reason for this uncertainty is that world classes are not tangible three-dimensional entities, but mental concepts, i.e. they ‘exist’ only in our minds.” (Aarts & Haegeman: 2006, 117)
According to Bußmann (Bußmann: 2002, 750 f.), parts of speech emerge from a process of classification which takes into account formal and functional features of the words to be classified as parts of speech. Aarts and Haegeman point out that these parts of speech or word classes can be viewed as “[…] abstractions over sets of words displaying some common property or properties.” (Aarts & Haegeman: ibid.).
Bußmann (Bußmann: ibid.) further states that attempts to classify parts of speech go back to the ancient world of Plato and Aristotle. Plato developed a system which split up speech acts into two categories, namely Onoma (name) and Rhema (declaration). Together, these categories form the logos, the speech or discourse.
(1.) The earth is flat.
In example (1.), The Earth would be the onoma, while is flat is the rhema of the sentence. Other related labels for onoma and rhema would be noun, subject, np, thing or argument for onoma and verb, predicate, vp, property or function for rhema.
Aristotle later added the category of syndesmos, which are those things that connect onoma and rhema into sentences. In traditional grammatical terminology, the category syndesmos might include such word classes as conjunctions, prepositions, pronouns and articles. But while Plato only made a logical division between parts of speech in terms of their discourse and meaning, Aristotle’s new category is grounded on a formal level. Aristotle hereby set an example for many others to follow; by adding more and more formal features to a model initially based solely on logical criteria, two different levels of categorisation were mixed which, under close scrutiny, turned out to not have been well combined indeed.
In the first century B.C., the grammarian Dionysios Trax developed a theory of eight parts of speech, which is still the basis of current attempts to classify parts of speech. This model includes nouns, verbs, adjectives, articles, pronouns, prepositions, adverbs and conjunctions. However, Bußmann (ibid.) shows that there exist a number of difficulties in fitting various words into these categories. First of all, the different consideration or accentuation of those aspects that might be employed for grouping different items into the parts of speech categories mentioned result in different categorisations or group memberships of these items. Bußmann lists morphological aspects (e.g. differentiation between inflecting and non-inflecting words), syntactic aspects (e.g. the ability to occur in certain phrases like noun phrase or verb phrase, the ability to occur with an article) and semantic aspects, which include the observation that the parts of speech such as noun, adjective and verb are based on logical categories (‘substance’, ‘property’, ‘process’) while conjunctions and prepositions are based on the category of ‘relation’. The difficulties that arise have their roots in the inconsistency in which the aforementioned criteria are applied to classify parts of speech. These criteria are often arbitrary or overlap, like in the case of numerals. Numerals on the basis of their common lexical features (they signify numbers and masses) form their own class, even though their single members, from a syntactic point of view, might behave like nouns (thousands of people), adjectives (one book) or adverbs (he called thrice) (examples taken from Bußmann, ibid., translations by me). Another difficulty arises from homonymy, which occurs when two or more words that sound the same have to be included within different parts of speech, e.g. rain (verb) and rain (noun) (my examples).
Concerning the semantic aspects, Aarts and Haegeman (Aarts and Haegeman: 2006, 118) point out that the so-called notional definitions which might be useful in certain pedagogical settings (like school grammars) are not adequate for defining parts of speech. Notional definitions make statements like “a noun is defined as ‘a word that denotes a person [sic] place or thing’, and a verb is an ‘action word.’” (Aarts and Haegeman: ibid.). But when we look at a verb like to seem, we can clearly see that it does not denote an action, while nouns like action and activity certainly do not denote persons, places or things.
Different theories of grammar have different approaches for categorizing which words belong to which parts of speech. One of them is generative grammar. Here, classification is exercised according to distributional criteria:
“Alle sprachlichen Einheiten, die in einem Satz für die gleiche lexikalische Konstituente eingesetzt werden können, gehören zur gleichen Kategorie.“ (Bußmann: 2002, 751)
Distributional criteria might be used to supplement or even to replace meaning-based definitions of parts of speech (Aarts & Haegeman: ibid.). A frame like ‘determinative-adjective-X’ is a distributional test to determine whether the words in the position of X in the following two examples are nouns or not (examples taken from Aarts & Haegeman: ibid., emphasis by me).
(2.) a beautiful cat
(3.) * the beautiful cheerfully
This distributional test proves cat to be a noun, since it can occur in the string determinative-adjective-X, while cheerfully cannot be a noun for the opposite reason. In 3.1., I will point out why Croft, who designed a variant of Construction Grammar called Radical Construction Grammar, argues against the distributional method concerning its ability to determine parts of speech.
One does not have delve deeply into the workings of the distributional method as means of defining the parts of speech to find that the results it produces are often far from clear cut in terms of the class membership of certain words. The two related issues of multiple word class membership and word class boundaries also pose difficulties for defining parts of speech, as the following distributions show:
- Quote paper
- Johannes Huhmann (Author), 2007, Parts of speech in Construction Grammar , Munich, GRIN Verlag, https://www.grin.com/document/87305