Analogy

The term “analogy” as used in linguistics has a specific and a general meaning. In historical linguistics it usually applies to morphological change, more specifically, change within morphological paradigms; that is what this chapter is about. The broader use of the term applies in syntax and refers to a process by which innovative expressions are based on existing expressions rather than on rules, a point we return to in Chapter 8.

This chapter is concerned with changes in morphological form (rather than meaning, which we get to in the next two chapters). The definition of morphological analogy I will use is the following: the re-making of a word based on similarity to other existing words in the language. In the preceding chapter we saw that sound change introduces alternations and irregularities into morphological paradigms. Analogical changes are often a response to these alternations, either working to eliminate them or, less often, to extend them to new lexical items.

Analogical change is rarely lexically regular the way sound change often is. That is, while sound change tends to affect all the lexical items that have the appropriate phonetic conditions for the change, analogical change clearly works one item at a time and most commonly does not affect all lexical items or paradigms that have the requisite conditions.

Introduction

Grammaticalization is interesting enough as a language-specific phenomenon but it becomes all the more important and interesting with the discovery that very similar paths of semantic change can be documented in unrelated languages. Indeed, large-scale crosslinguistic studies of grammaticalization reveal that for each gram type there are only a few possible lexical sources (Bybee et al. 1994; Heine et al. 1991b; Heine and Kuteva 2002). In Chapter 6 we saw that future grams can develop from verbs or constructions with meanings of volition or obligation, from constructions signaling movement towards a goal, and in a few cases from temporal adverbs, such as Tok Pisin baimbai. These four sources account for the vast majority of future grams whose sources are known.

The consequence of these facts is that in all languages at all times the same mechanisms of change are operating on very similar lexical and grammatical material and thus producing similar results. This is not to say that grammaticalization is exactly the same across languages. Poplack 2011 compares the grammaticalization over two centuries of the movement futures in French, Portuguese, and Spanish and demonstrates that not only are the rates of change different, but the contextual factors that influence the use of the movement future are different across these languages. What this means in terms of the universality of grammaticalization paths is that the conceptual sources are similar (they are probably never exactly the same), and the mechanisms of change such as bleaching and inference are very similar across languages, but in each language there will be some specific conditions that push the long-term development more in some contexts than in others. These language-specific factors may be the existence of other, competing forms, the attachment of social meaning to one variant or another, or particular patterns of usage that skew distributions. Despite these language-specific factors, broad patterns of change that are similar across unrelated languages can still be identified.

Introduction

This book examines the topic of how and why languages change. This field of study has traditionally been called “historical linguistics” and under that label the history of particular languages has been studied, and methods for the comparison of languages and reconstruction of their family relations have been developed. While this book covers many of the traditional topics in historical linguistics, I have chosen to focus on the topic of how and why languages change because linguistic researchers see now more than ever before that language change is not a phenomenon of the distant past, but is just as evident currently in ongoing changes as it is when we look back into documents that show older stages of languages. Moreover, it has become clear that language change helps us explain the features of language structure because it provides a window onto how those structures come into being and evolve. Thus we identify explanations for the characteristics that language has by examining how language changes.

What we will see as we progress through the types of language change is that change is built into the way language is used. The mental processes that are in play when speakers and listeners communicate are the main causes of change. This helps us explain another very important fact: all languages change in the same ways. Since language users the world over have the same mental processes to work with and they use communication for the same or very similar ends, the changes that come about in languages from Alaska to Zambia fit into the same categories as changes found in English and French.

In this final chapter, we take a look at the possible sources or causes of linguistic change. Of course, we have discussed the causes of change throughout the book, but in this chapter we address the issue of causes more directly and more generally. In broad outline, we consider factors internal to language as causes of change in contrast to external causes, in particular the influence of other languages. In this way, Section 11.1 on internal sources contrasts with Sections 11.2 and 11.3, which deal with situations of languages in contact. In Section 11.1 we will look briefly at some theoretical approaches that address issues of diachronic change, in particular Naturalness Theory in Section 11.1.1 and generative theories in Section 11.1.2. In Section 11.1.3 we consider briefly whether child language is a plausible source of linguistic change. After the sections on language contact, in Section 11.4 we consider the idea that language is a complex adaptive system, in which dynamic factors inherent in the speaker, listener, and context produce change.

Internal sources: language use

This section will first review the approach presented in the previous chapters, drawing together some common mechanisms and general patterns that have been identified across the different types of change. The approach taken here recognizes the role of language use in creating and propagating change and conforms to usage-based theory so we will call it the usage-based approach. Next, the discussion turns to other proposals, in particular that coming from Naturalness Theory, which proposes that language structures change to become more natural or less marked. Finally, we consider the generative approach, which hypothesizes that language change occurs in the language acquisition process.

Introduction

Syntactic change will be treated here as consisting of changes in syntactic constructions, including the creation of new constructions and changes in constructions once they come into existence. The term construction will be used here both in its very traditional use, where one speaks of “the passive construction” or a “relative clause construction” and also in the way it is currently used in syntactic analysis by linguists such as Goldberg, for whom a construction is a form–meaning relation (Goldberg 1995, 2006). For Goldberg, words and morphemes as well as phrases and syntactic patterns are form–meaning mappings, but in this chapter, we will consider syntactic patterns that are conventionalized. Such patterns have slots or positions within them that can be filled by a range of different words or morphemes and are thus schematic. So, for example, a prepositional phrase is a construction consisting of a preposition and an NP object. Both positions are schematic in the sense that many different items can occur in each slot. The preposition slot, say in English, is schematic since any preposition can occur in it. The NP slot is even more schematic because the range of possible NPs in English is enormous. Viewing a construction as a form–meaning mapping means that we take constructions to convey an overall meaning that goes beyond just the meaning taken from the words and morphemes that comprise them.

The phenomena treated under syntactic change are intertwined with grammaticalization in two ways. First, many constructions have specific grammatical morphemes in them and they have developed by the process of grammaticalization. Thus, all the prepositions found in English (in, to, of, behind, below, after, and so on) have undergone grammaticalization giving rise to new preposition constructions. When a unit or set of units changes category, we can speak of syntactic change. For instance, if a verb in construction with another verb becomes an auxiliary, not only has grammaticalization taken place, but syntactic change has also occurred. Second, the creation of new constructions is driven by some of the same processes that drive grammaticalization: chunking, category expansion, generalization, and inferencing. Thus many of the themes of this chapter are familiar from the last two chapters.

Introduction

The last chapter discussed a large of set of common sound changes, and in this chapter we take up some of the consequences of sound change, some other kinds of common sound changes, and also some phonological changes that do not exactly fit the definition of sound change. We begin by examining the consequences of the sound changes discussed in the last chapter for phonemic systems, considering splits that create new phonemes and mergers that spell the demise of some old phonemes. Then we consider how phonemic contrast operates in chain shifts, taking up vowel shifts which can operate in a chain-like manner, just as the consonant shifts can, as discussed in the last chapter. We also consider how contrastive tone arises in a language and how, in languages with a stress accent, prosodic change is associated with a cluster of other changes. Then we consider certain types of dissimilation, metathesis, and change motivated by phonotactic patterns. Finally, we consider once again the causes of sound change and what factors can be considered to discover the causes of sound change.

Phonologization

As we discussed particular types of sound changes in the last chapter we noted that each type occurs independently in unrelated languages and further that the changes have a phonetic motivation, usually in articulation. Many linguists have observed that sound changes arise from the universal tendencies that are present in many if not all languages (Hyman 1975; Ohala 2003), but for such a universal tendency to become a sound change, it must take on language-specific properties. The process by which a universal phonetic tendency becomes a sound change and is incorporated into a language-specific phonological system is called phonologization (Hyman 1975). The first step in phonologization is for the phonetic tendency to extend beyond the normal range predicted by universal co-articulation effects. For example, while it is normal for a vowel before a nasal consonant in the same syllable to have some nasalization on it, when this nasalization comes to involve full velum opening for all or most of the duration of the vowel we would say that vowel nasalization is phonologized. It is important to note, however, that since this process is quite gradual, it may be impossible to determine the exact moment at which a phonetic tendency crosses over into the phonology of a language.

Introduction

In this chapter and the next we examine the process of grammaticalization (sometimes also called “grammaticization”), the process by which new grammatical morphemes come into being. As mentioned before, grammatical morphemes (= grams) can be contrasted with lexical morphemes (such as nouns and verbs) in that lexical morphemes are considered “open class” items, meaning a language can easily add new members to the class, while grams are considered “closed class” morphemes because once a class is established it does not easily add new members. As our discussion unfolds it will become clear why this is.

Examples of grammatical morphemes are affixes, auxiliaries, articles, pronouns, prepositions, and postpositions. All grammatical morphemes are very restricted in where they can occur – affixes are attached to stems, auxiliaries occur with main verbs, articles with nouns, pronouns in place of noun phrases, prepositions and postpositions with noun phrases. That is, grams occur in particular constructions. Part of the story of how they develop concerns how the constructions they are in develop.

In this chapter and the next we will discuss how these grammatical forms develop over time. You might be surprised to learn that they almost all come from lexical items, such as nouns and verbs, or combinations of lexical and grammatical items.

Introduction

In this chapter we consider how the words of a language change. In previous chapters we have considered many changes that affect words. In Chapters 2 and 3 the discussion of sound change showed that the phonetic form of words can change, usually in a regular manner. In Chapters 4 and 5 we saw that the morphological composition of words can change as well. In Chapters 6 and 7 we saw that some words in specific constructions undergo grammaticalization and become grammatical morphemes. In Chapter 8 we saw that the particular constructions that words occur in can also change, which can affect the meaning of the words.

Here we consider additional types of change in words or the lexicon. First, we discuss the ways in which languages acquire new words in Section 9.2. Then we consider how words change their meanings, spreading this discussion over several sections: first in Section 9.3 we discuss the nature of the categories that constitute word meaning, and then we treat the mechanism by which words change their meaning over time. In Section 9.4 we discuss some general tendencies in lexical semantic change, and then in Section 9.5 we discuss the way words that are formed by derivational affixes lose their compositional meaning. Finally, Section 9.6 adds a few comments about how old words are lost from a language.