standard_of_comparison.Rmd

---
title: "Standard of comparison in comparative constructions"
author: "Chiara Naccarato"
output: html_document

---
---------------------------------------

# Chapter
## Standard of comparison

Comparative constructions typically involve three elements: a predicate and two noun phrases denoting the object of comparison and the **standard of comparison**. In example (@one) the object of comparison is constituted by the first noun phrase (the tree), while the second noun phrase (the house) is the standard of comparison (the preposition than is the standard marker).

(@one) _The tree is higher **than the house**_.

This chapter focuses on the morphological coding of the standard of comparison in comparative constructions. In the languages of Daghestan, the comparative meaning of a construction does not involve inflecting the adjective by degree (i.e. there is no morphological comparative degree), but the standard of comparison in such constructions is expressed by a **spatial form**, i.e. an inflected form of a nominal normally expressing a spatial relation. This is one of the most common strategies used to mark the standard of comparison in the languages of the world ([see](https://wals.info/feature/121A#4/41.57/33.00)).

Cross-linguistically, other common strategies involve the use of particles (other than those expressing spatial relations), predicates meaning ‘exceed, surpass’, and conjoined comparatives (see Stassen 2013). Sometimes more than one strategy is available in one and the same language. In the languages of Daghestan, for instance, constructions with postpositions are also found. 

Depending on the nature of the marker employed, spatially coded comparatives can be divided into three types: At-comparatives (i.e. essives), To-comparatives (i.e. latives), and From-comparatives (i.e. elatives) (cf. Stassen 2013). 

This variation within the group of spatially coded comparatives, which was not considered for the purposes of the WALS map of comparative constructions, is the focus of the present investigation. Our goal is to classify the languages of Daghestan according to the **type of spatial form used to mark the standard of comparison**.

## Spatial case systems in the languages of Daghestan

Most Daghestanian languages feature bimorphemic spatial case systems including:

- the common distinction between location (essive), destination (lative), source (elative), and path (translative)
- a specification of the configuration of an object in space (i.e. in a container, on a surface, behind another object, etc.). Markers conveying such meanings are often called **localization markers** (cf. Daniel & Ganenkov 2009: 674)

The Godoberi (Avar-Andic) examples in (@two) and (@three) demonstrate how spatial forms are constructed.


(@two) Godoberi (Saidova 1973: 70)

<ul>
<table> 
  <tr>
    <td><i>den&nbsp;</i></td>    <td><i>bel-qi&nbsp;</i></td>    <td><i>w-un-i-sːu&nbsp;</i></td>
  </tr>
  <tr>
    <td>1sg&nbsp;</td>    <td>mountain-ad&nbsp;</td>    <td>m-go-is-fut.def&nbsp;</td>
  </tr>
  <tr>
    <td colspan="3">'I will go to the mountains.'</td>
  </tr>
</table>
</ul>

(@three) Godoberi (Saidova 1973: 70)  

<ul>
<table> 
  <tr>
    <td><i>riχu-ƛi&nbsp;</i></td>    <td><i>reqi-me&nbsp;</i></td>    <td><i><b>bel-qi-ru&nbsp;</b></i></td>
  </tr>
  <tr>
    <td>sheep-gen&nbsp;</td>    <td>herd-pl&nbsp;</td>    <td>mountain-ad-el&nbsp;</td>
  </tr>
  <tr>
    <td><i>hiƛ’i&nbsp;</i></td>    <td><i>r-a-aƛa-da&nbsp;</i></td>    <td></td>
  </tr>
  <tr>
    <td>down&nbsp;</td>    <td>nhum.pl-come-prog.cvb-cop&nbsp;</td>    <td></td>
  </tr>
  <tr>
    <td colspan="3">'The herds are coming down from the mountains.'</td>
  </tr>
</table>
</ul>

## The standard of comparison in the languages of Daghestan

In the languages of Daghestan the standard of comparison is usually marked with a spatial form (see [Spatial forms]). There are cases, though less frequent, of languages featuring a specialized comparative marker (see [Specialized comparative markers]).

### Spatial forms

In comparative constructions, most languages of Daghestan mark the standard of comparison with a spatial form. In different languages, variation is observed not only with respect to the choice of a specific directionality marker (i.e. essive/lative/elative/translative), but also in terms of the localization marker employed, cf. examples (@four) and (@five), contrasting Standard Avar and Tindi (Avar-Andic), in which **superelative** and **contessive** are used respectively.

(@four) Standard Avar (Bokarev 1949: 165)

<ul>
<table> 
  <tr>
    <td><i><b>di-da-sa&nbsp;</b></i></td>    <td><i>ɬik’-a-w&nbsp;</i></td>    <td><i>qazaq&nbsp;</i></td>
  </tr>
  <tr>
    <td>1sg-sup-el&nbsp;</td>    <td>good-adjz-m&nbsp;</td>    <td>worker&nbsp;</td>
  </tr>
  <tr>
    <td><i>du-je=gi&nbsp;</i></td>    <td><i>šːʷ-ela-r=in&nbsp;</i></td>    <td></td>
  </tr>
  <tr>
    <td>2sg-dat=add&nbsp;</td>    <td>get-fut-neg=emph&nbsp;</td>    <td></td>
  </tr>
  <tr>
    <td colspan="3">'You will also not get a better farm worker than me.'</td>
  </tr>
</table>
</ul>

(@five) Tindi (Magomedova 2012: 79)

<ul>
<table> 
  <tr>
    <td><i>wacːi&nbsp;</i></td>    <td><i>kj'e-ja&nbsp;</i></td>    <td><i>rehã-ɬːiː &nbsp;</i></td>
  </tr>
  <tr>
    <td>brother&nbsp;</td>    <td>two-num&nbsp;</td>    <td>year-nm.obl.erg&nbsp;</td>
  </tr>
  <tr>
    <td><i>muk'u-w&nbsp;</i></td>    <td><i>ija&nbsp;</i></td>    <td><i><b>jacːu-č'i&nbsp;</b></i></td>
  </tr>
  <tr>
    <td>little-m&nbsp;</td>    <td>cop&nbsp;</td>    <td>sister-cont&nbsp;</td>
  </tr>
  <tr>
    <td colspan="3">'The brother is two years younger than the sister.'</td>
  </tr>
</table>
</ul>

In most languages of Daghestan elatives are used:

- superelative in:
  * Standard Avar, Bagvalal and Karata (Avar-Andic)
  * Standard Lezgian, Aghul, Archi[^1], Rutul and Tsakhur (Lezgic)
  * Itsari (Dargwa)
  * Hinuq, Tsez and Xvarshi (Tsezic)
- contelative in Botlikh, Godoberi and Chamalal (Avar-Andic)
- adelative in Kryz (Lezgic)

To these languages we might add those in which **ablative** forms are used (@six): Udi (Lezgic), Kumyk, Nogai and Northern Azerbaijani (Turkic), and Armenian. Ablatives are commonly employed to mark the standard of comparison in the languages of the world (Creissels 2009: 624).

(@six) North Azerbaijani (Širaliev & Sevortjan 1971: 47)

<ul>
<table> 
  <tr>
    <td><i>Bakı&nbsp;</i></td>    <td><i><b>Kirovabad-dan&nbsp;</b></i></td>    <td><i>böyük-dür&nbsp;</i></td>
  </tr>
  <tr>
    <td>Baku&nbsp;</td>    <td>Kirovabad-abl&nbsp;</td>    <td>big-cop.3sg&nbsp;</td>
  </tr>
  <tr>
    <td colspan="3">'Baku is bigger than Kirovabad.'</td>
  </tr>
</table>
</ul>

**Essives** are also quite frequent in Daghestan:
- superessive in Bezhta (Tsezic), Akusha (= Standard Dargwa), Mehweb and Kubachi (Dargwa)
- contessive in Andi, Bagvalal (as an alternative to the superelative) and Tindi (Avar-Andic)
- inessive in Bats (Nakh)

Northern Akhvakh (Avar-Andic) features the supertranslative case, while in Hinuq (Tsezic) the so-called aloc-lative (i.e. animate location lative) is used alongside the superelative (Forker 2019). 

### Specialized comparative markers

In some languages the standard of comparison is marked with a dedicated suffix, sometimes called “**comparative suffix**” (@seven). This is true for: Chechen and Ingush (Nakh), Hunzib (Tsezic), Archi, Budukh, Mukhad Rutul and Tabasaran (Lezgic), Khinalug and Lak. 

(@seven) Chechen (Nichols 1994: 30)

<ul>
<table> 
  <tr>
    <td><i>iza&nbsp;</i></td>    <td><i><b>suo-l&nbsp;</b></i></td>    <td><i>dika&nbsp;</i></td>    <td><i>v-u&nbsp;</i></td>
  </tr>
  <tr>
    <td>3sg&nbsp;</td>    <td>1sg-cmpr&nbsp;</td>    <td>good&nbsp;</td>    <td>m-cop&nbsp;</td>
  </tr>
  <tr>
    <td colspan="3">'He is better than me.'</td>
  </tr>
</table>
</ul>

However, it should be pointed out that, in some cases, this comparative suffix seems to include a spatial suffix, e.g. Khinalug *-q’il:i* (elative *-l:i*), Northern Tabasaran -t’an and Southern Tabasaran *-č’an* / *-dakan* (elative *-an*), Budukh *-wor* (elative *-r*). Archi *-χur* is reported to have some residual spatial usages (Daniel & Ganenkov 2009: 673-674).

## The “localization” issue

A quite controversial question, which lies behind the classification of spatially coded comparatives, concerns the semantics of localization markers. In Section 3 we grouped languages according to whether they feature a “super”, “cont” or “ad” marker (along with the distinction between elatives and essives). However, to my understanding, it is not straightforward what the semantic difference between these three labels is. This is also demonstrated by the fact that in the literature there is quite a great deal of contradictory information concerning the labeling of such markers. Moreover, in some cases, the decision to call a certain marker “super” or “cont” was taken by the author based on the information provided in the Russian-language literature, in which such labels are rarely employed (in most Russian-language grammars the different localization markers are listed by simply naming them “series 1”, “series 2”, and so on).

[^1]: The superelative suffix *-t:-iš* is mentioned by Mikailov (1967: 60) as the marker used to code the standard of comparison in Archi, but Kibrik (1977: 59) reports the dedicated comparative marker *-χur* instead.

```{r setup, include=FALSE} 
# make this part as a function and add it to example code
knitr::opts_chunk$set(echo = FALSE, message = FALSE)

# packages
library(tidyverse)
library(lingtypology)
library(RefManageR)

bib <- ReadBib(file = "./data/bibliography.bib")
BibOptions(style = "html", bib.style = "authoryear", cite.style = "authoryear", max.names = 2)

# load data
fe <- read_tsv("./data/features/tables/standard_of_comparison.csv") # feature dataset
```
```{r, child="_example.Rmd"}
```