Pernilla Hallonsten Halling

Results III: The construction level

1. Introduction

This chapter presents the results of examining the encoding of attr,
pred, and adv

in the languages of the sample on the level of the construction. In order to capture how the three functions are related in certain encoding patterns, it is necessary to analyze whole constructions. This analysis is based on the constructional-typological approach described in section 4.3, used as a method for consistent comparison of encoding where the scope is whole constructions. The most important results, presented in order of most to least common, are the following:

• 
  pred and adv, as opposed to attr, are very similarly encoded at the construction level in 13 of the languages (see section 7.3.1).
  
• 
  attr and pred, as opposed to adv, show identical encoding at the construction level in 12 languages (see section 7.2).
  

• 
  Nine languages have a construction that is intermediate between (see section 7.3.2).
  

pred
and adv

• 
  attr and adv, as opposed to pred, are encoded by the same type of construction in four languages (see section 7.4).
  

We may recall from chapters 4 and 6 that overlap refers to identical encoding in two or more functions. On the construction level, it is necessary to distinguish between total and partial overlaps. A total overlap means that the exact same encoding is used in more than one function. A partial overlap means that two encoding patterns are used in two functions, but that they overlap to a large extent in their structure. The construction- typological notation introduced in chapter 4 plays a key role in the analysis. The notation for English is therefore reproduced in (7.1),¹ with the actual examples included for illus- trative purposes. In the rest of the chapter, the Example line contains cross-references to examples in the text and appendix B.

(7.1) attr, pred, and adv in English

attr

Function: property modification within referring expression
Form: art ADJ N
Example: [The sad person] over there is Sheila

¹ Example (7.1) is a repetition of examples (4.9) and (4.16) in chapter 4.

preD

Function: property predication
Form: S be_cop ADJ Example: Sheila is sad

aDv

Function: property modification within predicating expression
Form 1: S V ADJ-ly _advz Form 2: S V ADJ_subset
Example: Sheila smiled sadly, Sheila runs fast

The notation in (7.1) illustrates that the same root, and in some cases the same lexeme, is used in all three functions. But such findings belong to the root and lexeme levels, as discussed in chapter 6. On the construction level, it is clear that the encoding is different for all three functions. Two constructions are nonetheless quite similar: those in pred and adv. One difference is whether the substantive copula be (pred) or the schematic verb slot (adv) is found. Another is whether the Adjective is used in its bare form (pred and Form 2 in adv), or the adverbializing ending -ly is required (pred). The constructional- typological notation does not tell us much here – there is no construction overlap in English – but the examples illustrate how the method is to be employed, highlighting differences as well as similarities.

In the following sections, overlaps are illustrated with examples from individual lan- guages, followed by constructional-typological notation. When an overlap of only two functions is discussed, only these two functions are included in the notation. Full nota- tion for examples from each language can be found in appendix B. As in the previous chapter, the overlap that they are argued to instantiate is indicated within square brackets in the right margin of the example heading (e.g. [attr pred]). The specific function is also indicated in the right margin of the example in question (e.g. attr). Some of the total overlaps are discussed in the respective sources as examples where the two functions cannot be distinguished, or where a construction is ambiguous. Other encoding overlaps are not mentioned in the sources, and are then based on my own conclusions. When I state that a certain language has an overlap of, e.g., attr and pred, this does not imply that it is the only encoding pattern attested in attr and pred: for some languages, this may be the case, whereas for others, the encoding overlap in question may coexist with other types of encoding. No claim is made here regarding what encoding pattern is more common. However, in many languages, encoding that does not overlap has been attested. The constructional-typological notation includes such encoding as well, with references to examples in appendix B. Accordingly, the phrasing that ‘a language has an overlap’ does not imply that it is the language as such that has overlapping functions. Rather, it is the constructions attested in specific languages that overlap. In most cases, construction-level overlaps imply overlaps at the root and lexeme levels as well (bearing in mind that an overlap of two functions on the lexeme level can be filled by items that instantiate the same or a larger overlap on the root level). But this is not necessarily the case, as will be illustrated specifically for the construction level overlap of attr and adv.

In the sections that follow, the overlaps are described in the same order as in chapter

6. First, the [attr pred] overlap is treated in section 7.2. This is followed by the [pred adv] overlap in section 7.3, which includes partial overlaps (section 7.3.1), constructions

in between
pred
and adv (section 7.3.2), and connections between constructions in

pred and adv (section 7.3.3). The [attr adv] overlap is discussed in section 7.4,
and potential overlaps of all three functions are treated in 7.5. For each overlap, a few examples of individual languages are discussed, followed by a constructional-typological summary of the examples in question. At the end of each section, the languages in which an overlap is attested are summarized in a table that includes constructional notation for each language.