Is Grammaticalization Unidirectional?

Seminar Paper, 2014

17 Pages, Grade: 1,7


Table of Contents

1. Introduction

2. Basic concepts and definitions in grammaticalization studies
2.1 Basic definitions and conceptual mechanisms
2.2 Lehmann’s parameters of grammaticalization
2.3 Defining degrammaticalization

3. Explaining the hypothesis of unidirectionality

4. Case studies
4.1 Pennsylvania Germanwotte: from modal auxiliary to lexical verb
4.2 The Englishs-genitive: from inflectional genitive suffix to a clitic

5. Conclusion


1. Introduction

Grammaticalization is a well-attested process of language change and presents a complex sub-field of linguistics. Although grammaticalization is believed to be a rather young area of linguistics, its history is as old as the history of linguistics (cf. Narrog & Heine 2011: 1). The term ‘grammaticalization’ itself was apparently first coined by the French linguist Meillet, a pioneer in the field of grammaticalization. In his work L’évolution des formes grammaticales (1912), Meillet describes the process of grammaticalization as “Le passage d’un mot autonome au rôle d’élément grammatical” (1912: 131 cited in Ferraresi 2014: 1) indicating a change of an erstwhile autonomous sign into a grammatical element. A more recent definition of grammaticalization is given by Hopper and Traugott who define it as “the process whereby lexical items and constructions come in certain linguistic contexts to serve grammatical functions, and, once grammaticalized, continue to develop new grammatical functions” (2008: xv).

Although the field of grammaticalization is already widely explored, its studies remain interesting since several of them have been the subject of critical discussions. One of the most interesting and challenging hypotheses in grammaticalization studies is presented by the unidirectionality hypothesis: “the claim that changes that fall into the category of grammaticalization always move into the direction – from more to less lexical or from less grammatical to more grammatical” (Börjars & Vincent 2011: 163).

However, is it not possible for a grammatical item to become less grammatical or even lexical? On the basis of Norde’s recent study on degrammaticalization (2009; 2012), the paper will take a closer look whether the unidirectionality hypothesis is entirely true or not.

The following paper is structured as follows: section two provides a brief overview of central concepts and definitions involved in grammaticalization and finally it presents Lehmann’s parameters of grammaticalization (1995). In section three, the paper features a central approach on how the unidirectionality hypothesis can be explained. Section four presents two valid counterexamples of the unidirectionality hypothesis with respect to Lehmann’s parameters (1995). To conclude, the paper summarizes the main results.

2. Basic concepts and definitions in grammaticalization studies

2.1 Basic definitions and conceptual mechanisms

The human language is a fairly complex and effective tool of information exchange. To understand the process of grammaticalization, we have to differentiate between two main classes: the lexical and functional items. Lexical items possess the primary meaning of a given sentence and they even have an own semantic value. In this context, we speak of categories such as nouns, verbs, adjectives and adverbs that can be identified as ‘lexemes’. Since these categories do have a concrete meaning, our mental lexicon is able to save their actual representations (cf. Ferraresi 2014: 1).

Functional elements such as articles, pronouns, conjunctions or prepositions serve as connective elements and thus have an abstract meaning enabling us to understand the sense of a given sentence (cf. Szczepaniak 2009: 1). Grammatical information can either be realized as a bound morpheme such as the plural indicator s in dogs or as a free morpheme indicating that it is not dependent on another word. The plural marker s such as in dogs is a so-called ‘grammeme’: the smallest unit of functional words (cf. Szczepaniak 2009: 2). Lexical items are in general realized as free morphemes whereas the process of grammaticalization is frequently accompanied by the change of a previous free morpheme to a bound one (cf. Ferraresi 2014: 4).

According to Hopper and Traugott, grammaticalization is the process whereby a lexical item becomes grammatical or a grammatical item becomes even more grammatical (2008: xv). Hence, the grammar is not a fixed and unchangeable system of rules, but a complex and dynamic entity that varies permanently. Hopper even speaks of a so-called ‘emergent grammar’: a grammar that constantly develops (Hopper 1987 cited in Szczepaniak: 2009: 5). Yet, what happens to a lexical item that develops to a grammatical element and which mechanisms are involved? The following example will clarify these questions from a diachronic perspective.

(Hopper & Traugott 2008: 2)

Regarding sentence (1), it becomes clear that the verb to go has lost its original status as a main verb and has developed into an auxiliary expressing immediate futurity. In other words, the main verb to go, which expresses any kind of motion, has lost its concrete meaning due to the processes of grammaticalization and thus received an abstract meaning. This change only appears in a very local context of specific directional constructions with non-finite complements indicating that the person is in fact leaving in order to do something in the future (cf. Hopper & Traugott 2008: 2).

The next mechanism involved in the process of grammaticalization presents the so-called ‘reanalysis’. Reanalysis, the process whereby the purposive be going (to…) is shifted to the auxiliary construction be going to, involves not only reanalysis of the be going to construction but also of the verb following it. Hence, the sentence [I am going [to marry Bill]] is rebracketed as [I [am going to] marry Bill] including a shift from progressive aspect to immediate future (cf. Hopper & Traugott 2008: 3).

As soon as the verb following the be going to construction is not compatible with a purposive meaning anymore, the reanalysis is discoverable such as in I am going to like Bill or It is probably going to rain all the time (cf. Hopper & Traugott 2008: 3). Be going to has been generalized and extended to contexts that were not available before. From a linguistic perspective, this phenomenon is called ‘analogical extension’.

Once the phenomenon of reanalysis has occurred, be going to undergoes some further modifications that are typical of auxiliaries, such as ‘phonological reduction’. As the phrasal boundary between –ing and to does not exist anymore, the three morphemes go-ing to are reduced into one morpheme (gonna). The present-day usage of the auxiliary be gonna is constrained by the original purposive meaning denoting the future of intention, plan or schedule. This kind of semantic reduction is known as ‘semantic bleaching’ (cf. Hopper& Traugott 2008: 3).

As the original meaning of the main verb to go has lost its status as an autonomous sign and developed as an obligatory future marker, we speak of ‘decategorialization’: the shift from a main verb to an auxiliary, which expresses immediate futurity.

Thus, grammaticalization presents a fairly complex process revealing a number of prototypical characteristics such as semantic bleaching, reanalysis, analogical extension, phonological reduction and decategorialization. The development of the future tense of be going to from a verb of motion is metaphorical in nature and presents only one of many examples in which the metaphorical transfer plays an important role: the transfer of a well-known concrete meaning to an abstract one. According to Heine et al. (1991: 47), there are certain cognitive categories leading to the phenomenon that we frequently tend to use concrete meanings in order to describe abstract circumstances. The future tense of be going to reveals a typical example in which the domain of space “is used as a metaphorical vehicle to refer to the domain of deictic time: the verb go to denoting physical action serves as a structural template for conceptualizing a grammatical notion, that is, deictic time” (Heine et al. 1991: 47). The concrete movement such as the be going to construction is “more easily grasped” (Heine et al. 1991: 47) than a concept falling into the category of abstract domains dealing with tense categories. Yet, how can we conclude that a lexical element has fully grammaticalized? In the following, the paper will provide a brief insight into Lehmann’s parameters of grammaticalization (1995): a set of criteria that measures the degree of grammaticalization of a sign.

2.2 Lehmann’s parameters of grammaticalization

Once a linguistic sign has been grammaticalized, it lost its status as an autonomous sign. To measure the degree to which a sign is grammaticalized, Lehmann distinguishes between three dimensions to determine its degree of autonomy accentuating three main aspects: semantic weight, cohesion and variability (1995: 122). As far as the process of grammaticalization is concerned, the semantic weight of a sign decreases. Cohesion, the inherent factor that a sign loses its autonomy because of its increasing relation to other signs, however, increases. Third, variability, the ability that a sign enjoys freedom of position within a sentence, decreases. These three aspects of grammaticalization are still rather abstract and difficult “to operationalize as analytic criteria” (Lehmann 1995: 122). Thus, Lehmann concretizes them with respect to a paradigmatic and a syntagmatic level. Paradigmatic means that a linguistic sign is chosen out of a group of several signs, whereas the syntagmatic level is concerned with particular combinations of linguistic signs (cf. Szczepaniak 2009: 20). Furthermore, he divides these levels into three separate categories. While the paradigmatic level is concerned with integrity, paradigmaticity and paradigmatic variability, the syntagmatic focuses on structural scope, bondedness and syntagmatic variability.

Table 1: The parameters of grammaticalization according to Lehmann(1995: 123)

illustration not visible in this excerpt

The paradigmatic weight or integrity of a sign refers to its possession of a certain semantic value, allowing it to maintain its own identity. Hence, main verbs such as to eat or to cook have a fairly high degree of integrity. Due to the so-called ‘erosion’ which consists of ‘desemanticization’ and ‘phonological attrition’, the degree of integrity of a sign decreases (cf. Lehmann 1995: 126; Szczepaniak 2009: 21). Whereas the term ‘desemanticization’ deals with the loss of semantic substance, the term ‘phonological attrition’ is concerned with the loss of phonological substance.

According to Lehmann, “paradigmatic cohesion or paradigmaticity is the formal and semantic integration of a paradigm as a whole and of a single subcategory into the paradigm of its generic category” (1995: 132). This implies that the members of a given paradigm are linked to each other “by clear-cut paradigmatic relations” (Lehmann 1995: 132). Correspondingly, there is an increase in cohesion of these members. This process is frequently accompanied by the shift from a major to a minor word class belonging to ‘decategorialization'. Therefore, a lexical item that has been finally reanalyzed as an auxiliary is by far more limited to a certain paradigm (cf. Lehmann 1995: 134).

The paradigmatic variability deals with the freedom by which the speaker chooses a sign (cf. Lehmann 1995: 137). Through ‘obligatorification’, the freedom of choice is reduced; thus the paradigmatic variability decreases. A typical example of the decrease of paradigmatic variability is presented by the German ‘Rezipientenpassiv’ in which the auxiliary bekommen such as in Das Auto bekommt einen Motor eingebaut has been reanalyzed as an auxiliary and can no longer be replaced by other verbs like the main verb erhalten (cf. Szczepaniak 2009: 5).


Excerpt out of 17 pages


Is Grammaticalization Unidirectional?
Johannes Gutenberg University Mainz
Seminar English Linguistics
Catalog Number
ISBN (eBook)
ISBN (Book)
File size
436 KB
Grammaticalization, Degrammaticalization, The English s-genitive, Pennsylvania German wotte, Unidirectionality
Quote paper
Kim Frintrop (Author), 2014, Is Grammaticalization Unidirectional?, Munich, GRIN Verlag,


  • No comments yet.
Read the ebook
Title: Is Grammaticalization Unidirectional?

Upload papers

Your term paper / thesis:

- Publication as eBook and book
- High royalties for the sales
- Completely free - with ISBN
- It only takes five minutes
- Every paper finds readers

Publish now - it's free