Explaining the irreversibility of grammaticalization

Term Paper (Advanced seminar), 2007

14 Pages, Grade: 1,0



1 Introduction

2 The irreversibility of grammaticalization
2.1 Grammaticalization as an invisible-hand process
2.1.1 Ecological conditions
2.1.2 Maxims of action
2.1.3 Invisible-hand process
2.2 Explaining irreversibility
2.2.1 Grammaticalization as an inflationary process

3 Criticism of Haspelmath’s theory

4 Extension of Haspelmath’s theory

5 Conclusion

6 References

1 Introduction

Grammaticalization, the process by which lexical categories turn in functional ones, has largely been recognized as being unidirectional (Cf. Haspelmath 1999: 1048). According to Haspelmath (1999), the irreversibility of grammaticalization can be explained within a usage-based theory of language change as proposed by Keller (1994), who sees language change as an invisible-hand process. This paper will give an account of Haspelmath’s explanation of how the grammaticalization process is embedded in this theory. In this context, special attention will be paid to how the irreversibility of grammaticalization is derived from one of the maxims of action that guide the speakers in their language use, namely the maxim of extravagance.

In order to make an evaluation of Haspelmath’s theory possible, two further views on it will be presented. The first is the critical reception by Geurts (2000a, 2000b), who objected to the notion of extravagance as being inappropriate and unnecessary for explaining the unidirectionality of grammaticalization. A closer look at his arguments will show whether this criticism is justified. Secondly, I will incorporate an extension of Haspelmath’s explanatory model as offered by Huelva Unternbäumen (2006). In this, the irreversibility of grammaticalization is not only derived from the maxim of extravagance, but also from three conditions that have to be fulfilled for grammaticalization and degrammaticalization to take place.

2 The irreversibility of grammaticalization

Supposing linguistic expressions were ordered along a continuum with lexical items at one end and functional ones at the other, grammaticalization refers to the process of language change that moves a linguistic item further towards the functional end of the continuum. This process comprises lexical categories turning into grammatical ones as well as grammatical items becoming even more grammatical. In a diachronic view, grammaticalization is identified to progress gradually and the six parameters of grammaticalization as described by Lehmann (1995) can be used to align the items along the continuum (Cf. Haspelmath 1999: 1044f.).

Whereas grammaticalization is involved in numerous instances of language change, the opposite, i.e. degrammaticalization[1], hardly seems to occur. Haspelmath estimates 99% of all shifts along the lexical/functional continuum to be grammaticalizations (Cf. ibid.: 1046). Since 1975, irreversibility has been acknowledged as an important feature of grammaticalization by all writers on the subject. Even linguists arguing against the unidirectionality claim recognize the asymmetry between grammaticalization and degrammaticalization (Cf. ibid.: 1047f.).

2.1 Grammaticalization as an invisible-hand process

As no satisfactory explanation for the irreversibility of grammaticalization has been given by earlier writers (Cf. ibid.: 1050ff.), Haspelmath employs Keller’s invisible-hand theory of language change (Keller 1994) to explicate how grammaticalization comes about and is influenced by performance factors, from which its irreversibility is then derived. Within this framework, language change is regarded as a unintended byproduct of the everyday language use of individual speakers. Rather than conceiving of language as an independent object it is traced back to individual linguistic acts (Cf. Haspelmath 1999: 1054). In this connection, language change is seen as an invisible-hand process,

that is, a phenomenon that is the result of human actions, although it is not the goal of hu- man intentions. An invisible-hand phenomenon is explained if it can be shown to be the causal consequence of individual actions that realize similar intentions (Ibid.).

In order to describe grammaticalization as an invisible-hand process, the ecological conditions and maxims of actions have to be explained first.

2.1.1 Ecological conditions

Haspelmath (Ibid.: 1054f.) identifies the three following ecological conditions within which the process of grammaticalization takes place:

a. Grammar as unconscious processing:

Linguistic units are ordered along a continuum from maximally free/con-scious/deliberate to maximally rule-bound/unconscious/automated. Items at the former pole are fully lexical elements, and items at the latter are fully func­tional (or grammatical) elements.

b. Basic discourse meanings:

Certain meanings of linguistic units are universally much more basic to speak­ing than others, i.e. they need to be conveyed much more frequently than others (e.g. ‘possession’, ‘instrument’ are more basic than ‘bicycle’ or ‘moon’ in this sense).

c. Frequency and routinization:

A general feature of cognitive processing in higher organisms is that frequent occurrence of a cognitive event leads to a greater ease of processing (routiniza­tion, automation), i.e. less attention is necessary to execute the same task.

It is important to see that the “ecological conditions are highly general properties of language and cognition that are independently motivated” (Ibid: 1056). Condition a draws a parallel between the earlier mentioned continuum between lexicon and grammar and modes of mental processing. Condition b is supported by the facts that the grammatical categories of different languages tend to overlap extensively and that most of the meanings of grammatical items come from the same small set of semantic categories. The final condition connects the ease of mental processing to the frequency of basic discourse meanings (Cf. ibid.).

2.1.2 Maxims of action

The five following maxims of action guide the speakers in their linguistic behaviour, the hypermaxim being the basis for the other four maxims:

1. Hypermaxim: talk in such a way that you are socially successful, at the lowest possible cost.
2. Clarity: talk in such a way that you are understood.
3. Economy: talk in such a way that you do not expend superfluous energy.
4. Conformity: talk like the others talk.
5. Extravagance: talk in such a way that you are noticed (Ibid.: 1055).

The speakers comply to these maxims during language performance in order to reach their goals. An important aspect is that these goals are not only of a communicative (i.e. being understood, talking economically), but also of a social nature. Without the hypermaxim driving speakers to social success, the maxim of extravagance, responsible for the use of innovative expressions, would not be justified. Since pragmatically salient innovations (e.g. metaphors) obscure the message and thus, make understanding more difficult, they would be eliminated by the clarity maxim. Additionally, the maxim of conformity would not allow other speakers to pick up new and rare expressions (Cf. ibid.: 1056). This, of course, implies that maxims 2-5 cannot all be followed simultaneously every time, but neither do they need to be as speakers do not always follow the same goals (Cf. Keller 1994: 140ff.).

If the goal is to achieve social success, being extravagant by using pragmatically salient innovations or, more specifically, by replacing functional elements with lexical elements, is one way of reaching it. The notion of extravagance is used in a generalized sense to describe the effect speakers intend to achieve with imaginative and vivid expressions, which is to be noticed by others (Cf. Haspelmath 1999: 1057).

2.1.3 Invisible-hand process

The ecological conditions and the maxims of action are needed to explain how language changes through the invisible-hand process. The starting point is an utterance of an individual speaker who wants to be noticed and therefore uses an innovative expression in place of a common one. The old element is necessarily replaced by a lexical expression because these are the only items that can be manipulated freely according to ecological condition a (Cf. ibid.). And if the lexical expression stands in the place of a functional item, “then this may trigger a grammaticalization process at the end of which the lexical item has turned into a grammatical item” (Ibid.).

Maxims 4 and 5 can cause other speakers to adopt the innovation, although this does not always have to happen. However, if they do adopt it, they show that they belong to the social group of the first speaker and, at the same time, they can distance themselves from speakers outside their social group by being extravagant themselves. If the social group using the innovation is socially influential, the new expression can spread throughout the linguistic community. With more and more speakers adopting it, the maxim of conformity will soon be reason enough to use the innovation.

If the new feature fulfils condition b, it will be used more frequently, which also entails increased predictability. Highly predictable items do not need to be pronounced very carefully because the danger of misunderstandings is low. So even with a slurred pronunciation speakers abide to both the maxims of clarity and economy.

According to ecological condition c, high frequency also brings about greater ease of cognitive processing. At this point, the lexical expression has become a functional item, that speakers can use with less conscious attention (Cf. ibid.: 1057f.). The reason for this is that “[r]outinization seems to be the explanation for several of the accompanying processes of grammaticalization” (Ibid.: 1058) such as obligatory use in certain contexts, fixed positions and phonological merging. At the end of the grammaticalization process, habituation causes the former new expression to no longer be perceived as pragmatically salient (Cf. ibid.: 1055).

2.2 Explaining irreversibility

For degrammaticalization to happen the starting point of the invisible-hand process would have to be a functional expression used in the place of a lexical one. According to condition a, grammatical elements are maximally rule-bound and processed automatically, i.e. unconsciously, in the speaker’s mind. Hence, if a speaker wanted to replace a lexical item with a grammatical one, “s/he would not be able to do this because functional elements cannot be used outside their proper places” (Ibid.: 1059).

But the crucial reason, in Haspelmath’s view, is that speakers do not have any motivation to create such an innovation as none of the maxims of action would justify such behaviour. To begin with, a grammatical element in the place of a lexical one would not be as salient or explicit as the latter. Consequently, such an innovation would not only violate the conformity maxim (which is violated by every innovation), but also the maxim of clarity (Cf. ibid.). Here, one could object that metaphors and innovations which lead to grammaticalization also violate these two maxims (Cf. section 2.1.2). But in those cases, the maxim of extravagance is sufficient to ensure compliance with the hypermaxim. Social success, however, cannot be achieved with innovations that form the starting point of degrammaticalization, therefore, the extravagance maxim would also be violated. The goal of such linguistic behaviour would be considerable understatement, not legitimized by any of the maxims. While the maxim of extravagance may lead to the use of a pragmatically salient innovation, the opposite of extravagance, which is conformity, does not lead to the opposite behaviour of using a functional item instead of a lexical one (Cf. Haspelmath 1999: 1059).

2.2.1 Grammaticalization as an inflationary process

The comparison of grammaticalization to inflation supports the significance of the extravagance maxim. As inflation is another instance of an invisible-hand process, parallels to the process of grammaticalization are not difficult to establish (Cf. ibid.: 1060f.). Haspelmath (Ibid.: 1061) cites the following example of inflation: “if a king buys the loyalty of some followers by making them into, say, Grand Dukes and the number of Grand Dukes in the country doubles, the value of that title is bound to decrease.” The king gains a short-term advantage (more loyalty), but as the value of the title decreases in the long run the Grand Dukes need a new incentive to remain loyal.


[1] Haspelmath (2004) as well as Huelva Unternbäumen (2006) distinguish between “degrammaticalization” and “antigrammaticalization”. While the former term stands for any kinds of (mostly abrupt) changes that involve a grammatical item losing its grammaticality, the latter is used for the reversal of a potential grammaticalization, showing the same intermediate stages as grammaticalization. Only antigrammaticalizations are exceptions to the unidirectionality claim. Since the main focus is on Haspelmath’s earlier works, I follow their use of “degrammaticalization” in the meaning of “antigrammaticalization” in this paper.

Excerpt out of 14 pages


Explaining the irreversibility of grammaticalization
University of Wuppertal  (Anglistik/Amerikanistik)
Catalog Number
ISBN (eBook)
ISBN (Book)
File size
418 KB
Explaining, Sprachwissenschaft, grammaticalization, irreversibility, invisible hand, rudi keller, keller
Quote paper
Eleni Stefanidou (Author), 2007, Explaining the irreversibility of grammaticalization, Munich, GRIN Verlag, https://www.grin.com/document/83227


  • No comments yet.
Look inside the ebook
Title: Explaining the irreversibility of grammaticalization

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